the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 17 Aug 2022
Local moisture recycling across the globe
Abstract. Changes in evaporation over land affect terrestrial precipitation via atmospheric moisture recycling and consequently freshwater availability. Although global moisture recycling at regional and continental scales are relatively well understood, the patterns and drivers of local moisture recycling remain unknown. For the first time, we calculate the local moisture recycling ratio (LMR), defined as the fraction of evaporated moisture that rains out within approximately 50 km from its source, and identify its drivers over land globally. We derive seasonal and annual LMR from multi-year (2008–2017) monthly averaged atmospheric moisture connections at a scale of 0.5° obtained from a Lagrangian atmospheric moisture tracking model. We find that, annually, on average 1.6 % of evaporated moisture returns as rainfall locally, but with large temporal and spatial variability, where LMR peaks in summer and over wet and mountainous regions. We identify wetness, orography, latitude, and convective available potential energy as drivers of LMR, indicating a crucial role for convection. Our results can be used to study impacts of evaporation changes on local precipitation, with widespread implications for, for example, regreening and water management.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(3038 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(3038 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
CC1: 'Comment on egusphere-2022-612', Ruud van der Ent, 24 Aug 2022
Dear authors,
I am happy to see that regional recycling at the global scale gets some renewed attention. However there are a couple of points I wish to bring to your attention:
- what's the reason to call your metric 'local' recycling? Isn't it just regional recycling for grid cells of 0.5 arcdegree x 0.5 arcdegree?
- I find the novelty somewhat overstated. As far as I understand the novelty is simply the fact that you calculate regional recycling on a higher resolution grid than other global studies, but the conceptual calculation is very old (see reference list in van der Ent and Savenije (2011) for example).
- a gridcell of 0.5 arcdegree in let's say Stockholm is two times smaller in area than a gridcell around the equator
- moreover, that same grid cell is 3 times smaller in length in east-west direction, and, therefore the dominant wind direction is rather influential on its value.
- in other words, the regional recycling metric or LMR is scale and shape dependent and as such its values cannot be compared from region to region.
- the search for a relation between LMR and other quantities that do not suffer from the scale and shape dependency (precipitation, evaporation, CAPE, biomes etc. ) is therefore fundamentally skewed.
- In Van der Ent and Savenije (2011) I had a suggested alternative metrics, which actually have local meaning for the recycling process, which are the local length scale of precipitation recycling and the local length scale of evaporation recycling. Surely these also rely on a few assumptions, but they do not suffer (or at least to a much more limited extent) from the scale and shape dependency. Please consider this approach or think of a better way to make your metrics scale and shape independent.
Citation: https://doi.org/10.5194/egusphere-2022-612-CC1 -
CC2: 'Reply on CC1', Ruud van der Ent, 24 Aug 2022
Correction, the length of a gridcell in Stockhom (at 59 deg N) is of course also around 2 times smaller than a gridcell a the equator and not 3 times, although further north gridcells that are 3 or more times smaller obviously do exist. Nonetheless the same arguments hold.
Citation: https://doi.org/10.5194/egusphere-2022-612-CC2 -
AC1: 'Reply on CC1', Jolanda Theeuwen, 16 Sep 2022
Dear Dr. van der Ent,
Thank you for your time and the comments you provided on our manuscript: ‘Local moisture recycling across the globe’. Your comments are of great value to improve our manuscript. Below we respond to each point separately to discuss how we will implement them.
what's the reason to call your metric 'local' recycling? Isn't it just regional recycling for grid cells of 0.5 arcdegree x 0.5 arcdegree?
Currently, in literature a similar concept (regional recycling) is used to describe recycling over areas with varying size. Such an area is often related to a peninsula, catchment, etc. Our metric is different from this as we provide information on a more specific and smaller scale namely 0.5 degrees. This is currently the best resolution available for such data with global coverage. We refer to our new metric as the local moisture recycling ratio, first, to highlight the difference with previous studies on regional recycling. Second, this resolution approaches a length scale where local processes have relevant contributions. Feedbacks between the land surface and atmosphere at this spatial scale are relevant for regreening projects. So, we decided to call our metric local recycling as we see potential to use it to study local feedbacks.
I find the novelty somewhat overstated. As far as I understand the novelty is simply the fact that you calculate regional recycling on a higher resolution grid than other global studies, but the conceptual calculation is very old (see reference list in van der Ent and Savenije (2011) for example).
We see there are some similarities to the work done by Van der Ent and Savenije (2011), yet we believe our work is novel. In our paper we assess the impact of the spatial scale within which the moisture recycles by calculating three different types of local moisture recycling. Furthermore, we assess potential drivers of local moisture recycling and by doing so we contribute to the physical understanding of local recycling. However, we will include the work from Van der Ent and Savenije (2011) in our introduction and we will discuss in more detail how our work compares to the work done by Van der Ent and Savenije. Furthermore, we will explain in more detail how our analyses add to the current knowledge, i.e., we can better understand the spatial patterns in local recycling by assessing its drivers. Additionally, the previous work was conducted using output from a different moisture tracking model (i.e., WAM2-layers), which assumes complete mixing of evaporated moisture within two atmospheric layers. For short time scales complete vertical mixing might not be realistic. The UTrack model distributes the evaporated moisture along the vertical moisture profile and therefore, might be more suitable for analyses on a smaller scale. However, we are thankful for this comment which helps us specifying the novelty of our research better.
The following points will be addressed after the last point.
A gridcell of 0.5 arcdegree in let’s say Stockholm is two times smaller in area than a gridcell around the equator
moreover, that same grid cell is 3 times smaller in length in east-west direction, and, therefore the dominant wind direction is rather influential on its value.
In other words, the regional recycling metric or LMR is scale and shape dependent and as such its values cannot be compared from region to region.
The search for a relation between LMR and other quantities that do not suffer from the scale and shape dependency (precipitation, evaporation, CAPE, biomes etc. ) is therefore fundamentally skewed.
In Van der Ent and Savenije (2011) I had a suggested alternative metrics, which actually have local meaning for the recycling process, which are the local length scale of precipitation recycling and the local length scale of evaporation recycling. Surely these also rely on a few assumptions, but they do not suffer (or at least to a much more limited extent) from the scale and shape dependency. Please consider this approach or think of a better way to make your metrics scale and shape independent.
We see that the spatial scale affects moisture recycling ratios, and we believe this is a very useful comment that will help us to improve the quality of our manuscript. To assess this effect for our results, we scaled the local moisture recycling ratio of each grid cell to an area of 50 km x 50 km (see supplement for Figure R1). The relatively large difference between local recycling and scaled local recycling at high latitudes indicates that the local moisture recycling is more uncertain for higher latitudes. For the rest of the globe, we find that the general patterns of local recycling we describe in our manuscript are consistent with and without scaling. The terrestrial surface at lower and mid latitudes are most important for moisture recycling and as the pattern of the scaled and non-scaled local recycling ratio are similar here, we believe the grid cell size causes only a small bias here.
Besides scaling, Van der Ent and Savenije (2011) presented another metric to describe the local moisture recycling, namely, the length scale of evaporation recycling (we will refer to this metric as length scale). They found that this metric scales with their definition of local recycling ratio and has a value typically in the order of 1000 km globally. We did calculate this length scale for our data (see supplement for Figure R2), and its patterns are similar to the patterns we found for local recycling with large values (i.e., small length scales) over tropics and mountainous regions and small values (i.e., large length scales) over desert areas. Similar to the result from Van der Ent and Savenije (2011) we find length scales in the order of 1000 km. Although it is an important metric, we believe that it is more difficult to apply it to, for instance, the impact of land use change on precipitation locally. For this, one needs to determine the amount of rain that recycles locally, and length scale does not quantify this whereas the local moisture recycling ratio does. Yet, to apply local moisture recycling locally, we believe it is important to better understand the local moisture recycling ratio first, and with our study we add to its understanding. In the discussion of our paper, we will discuss the differences between the local moisture recycling ratio and this length scale to allow readers to better assess what metric to use when addressing research questions related to moisture recycling as we do see value in using the length scale for research questions related to non-local effects.
Concerning the comment on the impact of wind direction, we fully agree that the dominant wind direction affects the value of LMR. We can imagine that especially on higher latitudes, where variation between grid cell shape and size is relatively large and the difference between the zonal and meridional length of the grid cell is large, wind direction might have a strong impact on local recycling as you indeed pointed out. We will clarify the impact of dominant wind direction in our discussion to create awareness of this effect amongst the readers.
Finally, you state that the analysis in which we study correlations between the local recycling and different variables is fundamentally skewed due to the scale and shape dependency, which some of the other variables don’t have. We agree that the set-up of our study may result in skewed output, which is intrinsic to our data. Therefore, there is no perfect method to make a comparison between recycling ratios in different regions to study the correlation between the local moisture recycling ratio and other variables. However, to address research questions related to quantifying local hydrology, the local moisture recycling ratio is useful and currently no other metric is available to quantify this globally. A comparison of the recycling ratio among different grid cells is difficult to physically interpret (Van der Ent and Savenije, 2011). Therefore, we need to gain a better understanding of the local moisture recycling ratio. We build this understanding by identifying some of its drivers. However, based on this comment we plan to account for this effect by conducting another analysis in which we classify the data based on latitude and calculate correlation coefficients for the data in these different classes. The grid cell sizes within each class will then be more comparable to minimize the skewness of the analysis. Furthermore, because mainly the higher latitudes are skewed, we excluded Antarctica from our analysis.
To summarize, we will use your comments to address the issue of scaling in more detail in our manuscript. This allows readers to put our results better in perspective. To support this, we will add figures R1 and R2 in the appendix of our manuscript. We would like to thank you for your constructive feedback as it is valuable for improving our manuscript.
On behalf of all authors,
Jolanda Theeuwen
-
CC3: 'Reply on AC1', Ruud van der Ent, 12 Oct 2022
The authors are doing a great job in giving prompt replies rather than waiting until after the public discussion, which illustrates that they're on top of things. I am generally satisfied with their responses, but would like to add a few constructive comments, which are attached in the supplement. I look forward to reading a final version of this manuscript when the formal review process is completed.
-
AC4: 'Reply on CC3', Jolanda Theeuwen, 18 Oct 2022
Dear Dr. Ruud van der Ent,
We are very thankful for your final comments, which are helpful. We will take them into account when we get the opportunity to revise our manuscript.
On behalf of all authors,
Jolanda Theeuwen
Citation: https://doi.org/10.5194/egusphere-2022-612-AC4
-
AC4: 'Reply on CC3', Jolanda Theeuwen, 18 Oct 2022
-
CC3: 'Reply on AC1', Ruud van der Ent, 12 Oct 2022
-
CC2: 'Reply on CC1', Ruud van der Ent, 24 Aug 2022
-
RC1: 'Comment on egusphere-2022-612', Anonymous Referee #1, 15 Sep 2022
Summary
Theeuwen et al. illustrate the local moisture recycling ratio (LMR) globally, i.e. the fraction of evaporation that rains out over the same region or its neighborhood. Using simulations from UTrack driven with ERA5 reanalysis from 2008-2017, they compare the LMR over different neighborhood sizes and illustrate its seasonality. To unravel the drivers of LMR, they calculate the Spearman correlation of the defined LMR to other variables, such as orography, latitude, convective available potential energy, and so on. The discussion centers around the use of these LMR estimates to guide land- and water management practices.
RecommendationThe manuscript is generally well written, but it is, in my opinion, missing novelty and/or a fresh perspective on moisture recycling indices that aim to guide land- and management strategies. Thus, the manuscript requires some major revisions and potentially a slightly different direction to make it a novel and interesting contribution. I will elaborate on my major concerns below.
Major points1. Novelty
The manuscript repeatedly claims that local moisture recycling ratios are calculated "for the first time" (l. 9) and that "it is unknown which fraction of moisture recycles within its source location, and how this recycling varies across the globe" (l. 22-23). However, this is not the first study to do exactly this: Van der Ent et al. (2010) and Van der Ent & Savenjie (2011) already featured such local evaporation recycling ratios and calculated them globally. Furthermore, 'evaporationsheds' (see e.g. Van der Ent & Savenije, 2013) contain the exact same information and papers and data sets have been published on this, see e.g. Link et al. (2020).Unfortunately, I also cannot consider the approach or the objective referred to in the discussion novel: the perspective on understanding the potential influence of land cover, and land- and water management practices via moisture recycling is not new either. Keys et al. (2016), for example, describe this in the context of 'ecosystem services' or 'water security' (Keys et al. 2020) - to name just a few examples. And this is also the subject of all 'green water' studies (e.g., te Wierik et al., 2021; te Wierik et al. 2020).
2. Moisture recycling drivers
I do, however, like the idea of looking at the drivers of moisture recycling; but the current analysis of the drivers is rather simple. In particular, I am a bit hesitant about the variables used to unravel the drivers of LMR, and the methodology used to do so. First of all, while I understand that there is a latitudinal dependence of moisture recycling, I wonder if 'latitude' is the real driver here. Shouldn't it rather be wind, incoming solar radiation and maybe even the underlying area of a grid cell (that differs with latitude)? Similarly, is it fair to use 'evaporation' and 'precipitation' as drivers of LMR? Isn't LMR defined based on these two fluxes? Of course, there is a dependency on both fluxes then... Second, calculating (globally averaged?) Spearman correlations to unravel drivers of LMR is a cheap way of doing this. LMR and any variable in Tab. 1 may be correlated through a third variable that represents the 'true' driver. Or in other words: a correlation does not imply causality.
3. Issues of scale
The definition of what is considered 'local' is rather random. The authors claim that the LMR is based on approx. 50km around the source; however, they also illustrate different definitions of this scale parameter, i.e. 1 grid cell, 9 grid cells and 25 grid cells. The argument for chosing 9 grid cells is rather vague: "To keep the spatial scale as small as possible but to still have a spatial pattern that we can explain physically" (l. 88-89). Could the authors explain why other patterns cannot be explained physically? Is there some lower limit to what the forcing and/or the model can represent? If so, could this limit be determined in a reasonable manner?Some suggestions
To make this a novel and interesting contribution in the field of moisture recycling, a bit more effort may be needed. The authors could, for example, compare their evaporation recycling ratios to the ones from Link et al. - I assume that much more could be learned from the difference of these data sets. Alternatively, the 'true' drivers of moisture recycling could be assessed, using a more sophisticated method to do so. Or the issue of scale and what can be considered local, given the spatio-temporal resolution of the forcing, could be put into focus... these are, however, just some suggestions that I could envision and that would make this paper novel and interesting to me. The authors do not need to follow those.
Minor points- l. 52-53: "Parcels are tracked for up to 30 days or up to the point at which only 1% of their original moisture is still present. " - can this be longer than 30 days?
- Eq. 1-3: this refers to different areas across the globe; where do the 50km from the abstract come in?
- Uncertainty of UTrack is not assessed at all; at least assumptions in the model should be summarised in the Methods section as well.
- Fig. 1: it should probably be "grid cell" and not "grids" in the subtitles
- l. 85-85: "These results seem to indicate that the tracking method we use is not sufficient to define recycling within one grid cell."; maybe it's not the tracking method but the (temporal) resolution of the forcing that is used, or the number of parcels tracked?
- l. 86-87: "Finally, scaling recycling to the number of grid cells, we find r9 and r25 do not relate linearly." Could you elaborate how you scaled this? P is not uniformely distributed across the 9 or 25 grid cells considered here, so I would not expect that there is a linear relationship?
- A suggestion: a uniform color scheme for Figs. 1-2 would be helpful
- l. 111-114: "Both convective and large-scale precipitation correlate with LMR (Table 1), however neither the fraction of convective precipitation nor the fraction of large-scale precipitation correlates with LMR (Table 1). Furthermore, evaporation correlates positively with LMR (Table 1, Fig. 3) indicating that the strong relation between P and LMR is not the only factor that causes a correlation between wetness and LMR." - does it make sense to correlate LMR with P? And as LMR is based on E and P, it needs to be correlated to E as well, right? E could also correlate with LMR because of P... there are so many dependencies here that it is difficult to unravel the real drivers.
- The relation between LMR and convection is not surprising; however, what would be novel was if large-scale and convective precipitation were tracked separately...
- l. 146ff: Are the correlations, especially with convective precipitation and large scale precipitation, subject to spatial and temporal scales?
- l. 155-174: discussion on biomes and deforestation a bit misplaced; not motivated in the introduction at all
- l. 179-181: well-mixed assumption is often hidden in many tracking studies; as far as I understand this is also the case for UTrack - and a recent study illustrated the impact using another Lagrangian model (Keune et al., 2022)
- l. 185ff: this should really be described in the methods, in my opinion
- l. 207: relation to agricultural water management remains unclear to me
- l. 234f: while I understand that you aim to use the LMR as a proxy for regions, in which land and water management may help foster moisture recycling, I don't think this scales at all. To assess the potential of the LMR as a proxy, it would be useful to know if, e.g. an increase in local E by say 10% also leads to an increase of local P by 10%. As you discuss correctly: there are many more factors that play a role here - not just the average recycling ratio; and I am missing an attempt to look at the 'true' drivers of LMR or at least an analysis that moves towards a better suited proxy to estimate the benefit/loss of water due to land- and water management practices in the conext of moisture recycling...
References- Keune, J., Schumacher, D. L., and Miralles, D. G. (2022), A unified framework to estimate the origins of atmospheric moisture and heat using Lagrangian models. Geosci. Model Dev., 15(5), 1875-1898.
- Keys, P. W., Porkka, M., Wang-Erlandsson, L., Fetzer, I., Gleeson, T., and Gordon, L. J. (2019), Invisible water security: Moisture recycling and water resilience, Water Security, 8, 100046.
- Keys, P. W., Wang-Erlandsson, L., and Gordon, L. J. (2016), Revealing invisible water: moisture recycling as an ecosystem service, PloS one, 11(3), e0151993.
- Link, A., van der Ent, R., Berger, M., Eisner, S., and Finkbeiner, M. (2020), The fate of land evaporation – a global dataset, Earth Syst. Sci. Data, 12, 1897–1912, doi:10.5194/essd-12-1897-2020.
- te Wierik, S. A., Cammeraat, E. L. H., Gupta, J., & Artzy-Randrup, Y. A. (2021), Reviewing the impact of land use and land-use change on moisture recycling and precipitation patterns, Water Resour. Res., 57, e2020WR029234. doi:10.1029/2020WR029234.
- te Wierik, S. A., Gupta, J., Cammeraat E. L. H., Artzy-Randrup, Y. A., (2020), The need for green and atmospheric water governance, WIREs Water, 7:e1406, doi:10.1002/wat2.1406.
- van der Ent, R. J., Savenije, H. H. G., Schaefli, B., and Steele-Dunne, S. C. (2010), Origin and fate of atmospheric moisture over continents, Water Resour. Res., 46, W09525, doi:10.1029/2010WR009127
- van der Ent, R. J., and Savenije, H. H. G. (2013), Oceanic sources of continental precipitation and the correlation with sea surface temperature, Water Resour. Res., 49, 3993– 4004, doi:10.1002/wrcr.20296.
- van der Ent, R. J. and Savenije, H. H. G. (2011), Length and time scales of atmospheric moisture recycling, Atmos. Chem. Phys., 11, 1853–1863, doi:10.5194/acp-11-1853-2011.
Citation: https://doi.org/10.5194/egusphere-2022-612-RC1 -
AC3: 'Reply on RC1', Jolanda Theeuwen, 06 Oct 2022
Dear reviewer,
Thank you for your time and for reviewing our manuscript: ‘Local moisture recycling across the globe’. We appreciate your feedback, which is very helpful to improve our manuscript. Below we respond to each major point of feedback separately to discuss how we will implement them. The minor comments will be addressed when we get the opportunity to revise our manuscript.
- Novelty
The manuscript repeatedly claims that local moisture recycling ratios are calculated "for the first time" (l. 9) and that "it is unknown which fraction of moisture recycles within its source location, and how this recycling varies across the globe" (l. 22-23). However, this is not the first study to do exactly this: Van der Ent et al. (2010) and Van der Ent & Savenjie (2011) already featured such local evaporation recycling ratios and calculated them globally. Furthermore, 'evaporationsheds' (see e.g. Van der Ent & Savenije, 2013) contain the exact same information and papers and data sets have been published on this, see e.g. Link et al. (2020).
Unfortunately, I also cannot consider the approach, or the objective referred to in the discussion novel: the perspective on understanding the potential influence of land cover, and land- and water management practices via moisture recycling is not new either. Keys et al. (2016), for example, describe this in the context of 'ecosystem services' or 'water security' (Keys et al. 2020) - to name just a few examples. And this is also the subject of all 'green water' studies (e.g., te Wierik et al., 2021; te Wierik et al. 2020).
We thank the reviewer for explaining why they believe our manuscript is not as novel as stated in the manuscript. Considering that the comment posted by Dr. Ruud van der Ent, the review posted by Dr. Patrick Keys and your review all include this point of feedback, we see the importance of improving on this point. Therefore, we will better acknowledge all relevant previous studies, for example the studies mentioned by the reviewer. Following this we will be better capable of highlighting the differences between our work and the previous work as we do believe there are important new steps being made in our paper. Namely, first, we study the effect of its spatial scale on local moisture recycling. of our definition of local, i.e., the area within which the moisture recycles, and second, we assess potential drivers of local moisture recycling. Besides, the datasets that were used to calculate local recycling in previous studies differ from the dataset used in our study. To obtain these datasets different models were used and the forcing data of these models has a different spatial resolution. We will mention this in our introduction and come back to it in our discussion in more detail by comparing it to the earlier work done by Van der Ent and Savenije (2011). Furthermore, we will highlight how our work compares but also deviates to the work done related to ‘ecosystem services’, ‘water security’ and ‘green water’ studies, to which the reviewer refers Those studies have a focus on source-sink relations in which the sink, apart from the source region, includes also remote locations. In contrast, our work aims to quantify and better understand local recycling. In contrast our work focusses on to quantify and better understand local recycling. The main reason is that previous research on atmospheric moisture connections mainly focusses on non-local water management even though research shows regreening can cause local drying. This suggests the relevance of studying the impact of land cover changes on the local water cycle. Local moisture recycling can help us here. Even though Van der Ent and Savenije (2011) calculated a similar type of recycling the link with preventing local drying has not been made yet, which we believe could be highly valuable. In addition, the spatial scale of 0.5 degrees allows better to study local impacts than the scale of 1.5 degrees. We are thankful for this comment as it helps us to specify the novelty of our manuscript better. We will do this by better acknowledging relevant previous studies.
- Moisture recycling drivers
I do, however, like the idea of looking at the drivers of moisture recycling; but the current analysis of the drivers is rather simple. In particular, I am a bit hesitant about the variables used to unravel the drivers of LMR, and the methodology used to do so. First of all, while I understand that there is a latitudinal dependence of moisture recycling, I wonder if 'latitude' is the real driver here. Shouldn't it rather be wind, incoming solar radiation and maybe even the underlying area of a grid cell (that differs with latitude)? Similarly, is it fair to use 'evaporation' and 'precipitation' as drivers of LMR? Isn't LMR defined based on these two fluxes? Of course, there is a dependency on both fluxes then... Second, calculating (globally averaged?) Spearman correlations to unravel drivers of LMR is a cheap way of doing this. LMR and any variable in Tab. 1 may be correlated through a third variable that represents the 'true' driver. Or in other words: a correlation does not imply causality.
We are happy that the reviewer likes the aim to understand the drivers of local moisture recycling but also understand that the reviewer likes to see more regional (latitudinal) tests that can potentially provide more understanding. In our study we aimed to identify non-linear relations between two variables as most processes cannot be properly described using linear relations and therefore, we used Spearman rank correlations. However, of course we agree that correlation does not imply causality and we will clarify this in the discussion of our manuscript. Moreover, we will use literature to discuss our findings from a mechanistic point of view (see also reviewer Keys). Furthermore, we agree that latitude is not the actual driver of moisture recycling, but that other variables, that correlate with latitude, drive local recycling. As such, we included latitude as a proxy for a combination of processes that have a strong latitudinal pattern. This was not properly described in our manuscript and therefore, we will clarify this in our revision.
Further we will also add more drivers to study the correlation between local moisture recycling and other potential drivers of local moisture recycling, such as solar radiation as suggested by the reviewer. We will not move to multiple regression models, as also clearly indicated by the reviewer it is true that many drivers themselves are correlated to each other. Therefore, we will keep the Spearman rank correlation test per driver. In addition, we plan to split the data in classes based on latitude to account and can then better understand how drivers can change per latitude class which can help to understand the causality. We believe adding more variables to our analysis will improve the understanding of the drivers of local moisture recycling and we are thankful for this comment.
- Issues of scale
The definition of what is considered 'local' is rather random. The authors claim that the LMR is based on approx. 50km around the source; however, they also illustrate different definitions of this scale parameter, i.e. 1 grid cell, 9 grid cells and 25 grid cells. The argument for chosing 9 grid cells is rather vague: "To keep the spatial scale as small as possible but to still have a spatial pattern that we can explain physically" (l. 88-89). Could the authors explain why other patterns cannot be explained physically? Is there some lower limit to what the forcing and/or the model can represent? If so, could this limit be determined in a reasonable manner?
We thank the reviewer for pointing out the unclarity concerning the spatial scale of local moisture recycling. We agree that the definition is partly arbitrary. We will mention this in our manuscript. Concerning the decision to use recycling over 9 grid cells, local recycling within one grid cell results in exceptionally low values over mountain peaks, yet not over all elevated terrain and relatively high values over the ocean. This pattern is inconsistent with the result found for recycling within 9 and 25 grid cells. The patterns for recycling over 9 and 25 grid cells can be explained as high values over mountains can result from convection as a result of orographic lift and relatively low values over ocean can be explained by the large atmospheric moisture transport due to strong winds. Possibly a numerical process is in place for the recycling within one grid cell, causing the pattern to be different from recycling over 9 and 25 grid cells. As the pattern of the latter two definitions are similar and agree with our understanding we decided to define local moisture recycling as the recycling over evaporated moisture within its source grid cell and its 8 surrounding grid cells. We will clarify this part in our results section. Furthermore, we will rephrase the sentence in which we state we cannot “physically” explain the pattern of recycling within one grid cell. We will omit the word physically and state we cannot fully explain the pattern of recycling within one grid cell. We believe these adjustments will clarify our decision and we thank the reviewer for pointing this out.
Some suggestions
To make this a novel and interesting contribution in the field of moisture recycling, a bit more effort may be needed. The authors could, for example, compare their evaporation recycling ratios to the ones from Link et al. - I assume that much more could be learned from the difference of these data sets. Alternatively, the 'true' drivers of moisture recycling could be assessed, using a more sophisticated method to do so. Or the issue of scale and what can be considered local, given the spatio-temporal resolution of the forcing, could be put into focus... these are, however, just some suggestions that I could envision and that would make this paper novel and interesting to me. The authors do not need to follow those.
We agree with these suggestions. As described before we will conduct more analyses by studying the relation between local moisture recycling and more potential drivers (e.g., solar radiation and other aspects of the energy balance). In addition, we will classify our data based and for each latitudinal class and conduct a spearman rank correlation analysis separately. Furthermore, we will compare our results to the dataset by Link et al. (2020). This means we need to upscale our findings to the 1.5 degrees resolution to match the dataset from Link et al (2020) to the dataset by Tuinenburg et al. (2020). This will give some insight into potential differences due to different use of models and simulated time periods, independent to the resolution used.
We would like to thank the reviewer for their constructive feedback as it is valuable for improving our manuscript. As mentioned before, in a future response will address the minor points of the reviewer.
On behalf of all authors,
Jolanda Theeuwen
References:
Link, A., van der Ent, R., Berger, M., Eisner, S., and Finkbeiner, M.: The fate of land evaporation – a global dataset, Earth Syst. Sci. Data, 12, 1897–1912, https://doi.org/10.5194/essd-12-1897-2020, 2020.
Van der Ent, R. J., & Savenije, H. H. G. (2011). Length and time scales of atmospheric moisture recycling. Atmospheric Chemistry and Physics, 11(5), 1853–1863. https://doi.org/10.5194/acp-11-1853-2011
Van der Ent, R. J., Savenije, H. H. G., Schaefli, B., & Steele-Dunne, S. C. (2010). Origin and fate of atmospheric moisture over continents. Water Resources Research, 46(9), W09525. https://doi.org/10.1029/2010WR009127
Keys, P. W., Wang-Erlandsson, L., and Gordon, L. J. (2016), Revealing invisible water: moisture recycling as an ecosystem service, PloS one, 11(3), e0151993.
te Wierik, S. A., Cammeraat, E. L. H., Gupta, J., & Artzy-Randrup, Y. A. (2021), Reviewing the impact of land use and land-use change on moisture recycling and precipitation patterns, Water Resour. Res., 57, e2020WR029234. doi:10.1029/2020WR029234.
te Wierik, S. A., Gupta, J., Cammeraat E. L. H., Artzy-Randrup, Y. A., (2020), The need for green and atmospheric water governance, WIREs Water, 7:e1406, doi:10.1002/wat2.1406.
Tuinenburg, Obbe A., Theeuwen, J. J. E., & Staal, A. (2020). High-resolution global atmospheric moisture connections from evaporation to precipitation. Earth System Science Data S, 12(4), 3177–3188. https://doi.org/10.5194/essd-12-3177-2020
Citation: https://doi.org/10.5194/egusphere-2022-612-AC3
-
RC2: 'Comment on egusphere-2022-612', Patrick Keys, 16 Sep 2022
SUMMARY
The authors of "Local moisture recycling across the globe" explore the concept of local moisture recycling (LMR), and its distribution and characteristics globally. The authors aim to understand the drivers of LMR and explore the heterogeneity of the phenomena across latitudes, biomes, and elevations. In general, I find the work interesting and the analyses mostly sound. I think that once the authors can make some changes — which I think would fall into the minor category, but I suppose could be major — this ought to be published.Here are several general comments followed by more line specific comments afterward.
GENERAL COMMENTS
1. "Drivers": I had some trouble with the word “drivers” being used to describe the role of e.g., CAPE, in LMR. While it's possible that some of the correlated variables could be causally related to LMR, there was no causal analysis completed (or a mechanistic explanation) as far as I could tell. Given that, it seems perfectly reasonable to say correlated phenomena (or similar) with regard to the variables that seemed to have convincing correlations with LMR, e.g., wetness or elevation.2. LMR Definition: This is purely a suggestion, but I would strongly encourage the authors to frame the LMR idea conceptually first (which I think would force the authors to argue more clearly for the novelty of the idea), and then provide the specific way that they define it in the article (e.g., "LMR is _________, which we define here as ______"). The reason being that LMR could be a conceptually useful idea on its own, but others may make entirely different quantifications, or indeed develop a more robust geophysical definition later on. By distinguishing your conceptual contribution from the physical definition, you may give the concept more scholarly and applied longevity. Again, just a suggestion.
3. Expand the Introduction: Given the intended scope of this work, the introduction should be expanded to include discussion of past moisture recycling analyses that relate to the LMR idea. There are quite a lot of moisture recycling studies that examine: local recycling (though perhaps not at the global scale), the scale dependence of moisture recycling, and the role of different types of vegetation in the scale dependence of moisture recycling (e.g., van der Ent et al., 2014).
4. Discussion of robustness of a 10-yr climatology: The current draft of the paper does not suitably discuss the appropriateness of the time scale associated with the Utrack dataset in the context of climate variation. Given that the authors are making claims about 'average rates', correlating with phenomena to determine their relationship (potentially causative) of LMR, etc., it is necessary to provide both a justification and a discussion (including of limitations) vis a vis a 10-year climatology. Given that there are inter-annual (e.g., ENSO, Indian Ocean Dipole) and decadal-scale (e.g., Pacific Decadal Oscillation), modes of climate variability that could be systematically affecting some of these results, it is necessary to explain the role of a 10-yr only analysis. Saying that the data are only available for 10 years is not really sufficient. Acknowledging this temporal limitation is critical also for making sure readers can appropriately interpret the results, which could change with a longer time series. Again — I recognize the authors have mentioned the topic of the length of record albeit briefly, but a more in-depth discussion is needed.
5. A 50-km definition: The authors select 50 km as the spatial scale of their LMR definition. A bit more discussion is needed to explain the logic and rationale of why such a blanket definition across the terrestrial surface is appropriate, and not an orographically-, latitudinally-, or biome-dependent definition. I’m not suggesting to change away from 50 km, but enough other research has found that moisture recycling ratios and spatial scales are associated with vegetation type, position on continent relative to prevailing winds, proximity to mountains, etc. that the authors need to support their 50km definition more strongly.
SPECIFIC COMMENTS
L10 Consider saying “defined here as…”L11 You could consider using the phrasing ”a 10-year climatology…”
L21 See General Comment about Expanding the Introduction.
L23 The sentence beginning with “However, it is unknown…” seems like it might not truly reflect the state of knowledge about moisture recycling globally, and overclaim the knowledge gap. You could consider situating this more concretely in what is understood (e.g., from the perspective of vegetation type having different length scaling associated with moisture recycling, etc.)
L38 The authors propose to explore “spatial-temporal variation across the globe” but a 10 year climatology is not enough to support the “temporal variation” component.
L67 What is the origin of these correlated variables? I could not see whether they came from ERA5 or somewhere else. Likewise, given the range in geophysical process scale of these variables, it would be good to comment here (or in the discussion of limitations) on the spatial scale of some of these, and whether some are more or less appropriate at the scales of analysis used in the study.
L67 See General Comment about the word “Drivers.”
L78 It is worth noting (either here, or in the Discussion) that the specific continental configuration (and mountain ranges) relative to prevailing winds and the associated biome composition matters quite a lot for determining the distribution of moisture recycling ratios. I know that the authors know this already (just looking at the author names and their past publications) so it is a noticeable gap in the logic.
Fig3 This is an interesting figure, but I think that the bottom left panel provides the most insight. I would recommend pulling this figure out on its own, so that it can be seen in a larger capacity, given the density of dots and information presented.
L135 Again, drivers is probably not quite right here, since there is no sufficiently causal mechanistic explanation of the link between e.g., CAPE and LMR.
L141 This seems like a perfect opportunity to situate the findings in the sweep of convective storm literature. I suspect that a convective storm meteorologist might not find the statement very surprising “our results suggest a positive relation between convection and LMR.” That doesn’t mean it shouldn’t be said — just that some references aligning this statement with the corresponding literature seem prudent.
L149 If I’m looking at the right Miyamoto 2013 reference, the argument appears to be that the number of convective features increases dramatically with resolution. That being said, the statement that “convection is a local scale process (i.e., spatial scale of 100 km)” might want to be adjusted.
L145 The entire paragraph needs better referencing, since the authors are making numerous claims regarding convective storms, and how the LMR analysis relates to that field. Greater referencing would also give me (the reader) more confidence that these claims are supported by the broader field.
L155 The global biome discussion is very interesting, but in its current presentation it both (a) reads as results, and (b) needs a supporting figure in the main text.
L168 These findings are very interesting, and are well discussed. I would use the density of supporting references and citations here as an example of what is necessary in the “convection” section above.
L203 The authors rely heavily (though not entirely) on Salmon et al 2011. Given the range of claims being discussed here, it seems prudent to include a few more references (than relying on Salmon et al. three times in the same paragraph.
L221 I encourage the authors to cite the work by Kirsten Findell in this paragraph (see Refs below), who provides a global analysis which blends empirical analysis and theory to explore how continental moisture recycling may change over the coming century.
L235 The sentence should be restructured for clarity.
L242 It might be interesting to be able to state the standard deviation associated with the 1.6% number that is quoted throughout the paper.
REFERENCES
Findell, K.L. et al. (2019) ‘Rising Temperatures Increase Importance of Oceanic Evaporation as a Source for Continental Precipitation’, Journal of Climate, 32(22), pp. 7713–7726.
van der Ent, R.J. et al. (2014) ‘Contrasting roles of interception and transpiration in the hydrological cycle – Part 2: Moisture recycling’, Earth System Dynamics, 5(2), pp. 471–489.
Citation: https://doi.org/10.5194/egusphere-2022-612-RC2 -
AC2: 'Reply on RC2', Jolanda Theeuwen, 04 Oct 2022
Dear Dr. Patrick Keys,
Thank you for taking the time to review our manuscript ‘Local moisture recycling across the globe’. We very much appreciate your feedback as we believe they will help us to improve our manuscript. Below we will shortly respond to each of the general comments you wrote in your review. We will address the specific comments and the general comments in more detail when we get the opportunity to revise our manuscript.
- "Drivers": I had some trouble with the word “drivers” being used to describe the role of e.g., CAPE, in LMR. While it's possible that some of the correlated variables could be causally related to LMR, there was no causal analysis completed (or a mechanistic explanation) as far as I could tell. Given that, it seems perfectly reasonable to say correlated phenomena (or similar) with regard to the variables that seemed to have convincing correlations with LMR, e.g., wetness or elevation.
We agree on the point that a correlation between LMR and any other variable does not imply causality in our study and therefore, we understand that “driver” might not be the best word to use in our manuscript. We thank you for this contribution; we will use different terminology throughout our manuscript and think that the term ‘factors’ is more appropriate. Furthermore, we will highlight in our discussion or methods section that a correlation does not imply causality in our study.
In addition we will better implement our hypothesis into our manuscript. We will explain what processes we expect LMR to be part of using previous literature. This will support our decision what variables we included in our study. For each variable in our study, we can also highlight whether we expect a direct relation between this variable and local moisture recycling or whether this variable is a proxy.
- LMR Definition: This is purely a suggestion, but I would strongly encourage the authors to frame the LMR idea conceptually first (which I think would force the authors to argue more clearly for the novelty of the idea), and then provide the specific way that they define it in the article (e.g., “LMR is _________, which we define here as ______”). The reason being that LMR could be a conceptually useful idea on its own, but others may make entirely different quantifications, or indeed develop a more robust geophysical definition later on. By distinguishing your conceptual contribution from the physical definition, you may give the concept more scholarly and applied longevity. Again, just a suggestion.
We very much appreciate this suggestion from the reviewer. We believe by framing LMR conceptually our proposal to study local hydrological impacts of land cover changes becomes clearer. This allows future studies to explore this novel concept of local moisture recycling using different definitions. We see how this indeed helps us to better highlight the novelty of our approach and we agree that by framing it as you suggest we better clarify the importance of moisture recycling locally regarding land cover change. Many thanks for this suggestion we will implement it in the introduction.
- Expand the Introduction: Given the intended scope of this work, the introduction should be expanded to include discussion of past moisture recycling analyses that relate to the LMR idea. There are quite a lot of moisture recycling studies that examine: local recycling (though perhaps not at the global scale), the scale dependence of moisture recycling, and the role of different types of vegetation in the scale dependence of moisture recycling (e.g., van der Ent et al., 2014).
To better highlight the novelty and aim we will expand the introduction and include more literature on studies related to moisture recycling. We will discuss the length scale of evaporation recycling in our introduction. We also will discuss its spatial patterns and explain its physical meaning. We will use this to clarify that the length scale is different from the local moisture recycling ratio and clearly show the added value of the local moisture recycling to the scientific community is. In our response to the community comment from dr. Ruud van der Ent we already did describe in more detail how we will implement the length scale in the introduction and discussion of our manuscript. Thank you for providing these examples.
- Discussion of robustness of a 10-yr climatology: The current draft of the paper does not suitably discuss the appropriateness of the time scale associated with the Utrack dataset in the context of climate variation. Given that the authors are making claims about ‘average rates’, correlating with phenomena to determine their relationship (potentially causative) of LMR, etc., it is necessary to provide both a justification and a discussion (including of limitations) vis a vis a 10-year climatology. Given that there are inter-annual (e.g., ENSO, Indian Ocean Dipole) and decadal-scale (e.g., Pacific Decadal Oscillation), modes of climate variability that could be systematically affecting some of these results, it is necessary to explain the role of a 10-yr only analysis. Saying that the data are only available for 10 years is not really sufficient. Acknowledging this temporal limitation is critical also for making sure readers can appropriately interpret the results, which could change with a longer time series. Again — I recognize the authors have mentioned the topic of the length of record albeit briefly, but a more in-depth discussion is needed.
We agree that the 10-year averaged data is not sufficient for a robust climatological analysis and will highlight this better in our manuscript. Our analysis, and the 10 years involved, are used for intra-variability (seasonlity) within the year. However, trends in multi-year climate variation could affect our results. To discuss this point, we will compare our results with the results from Link et al. (2020), as that study has a longer time period. To enable a comparison, we will calculate local moisture recycling ratios at a spatial scale of 1.5 degrees from both the dataset by Link et al. (2020) and the dataset by Tuinenburg et al. (2020).
- A 50-km definition: The authors select 50 km as the spatial scale of their LMR definition. A bit more discussion is needed to explain the logic and rationale of why such a blanket definition across the terrestrial surface is appropriate, and not an orographically-, latitudinally-, or biome-dependent definition. I’m not suggesting to change away from 50 km, but enough other research has found that moisture recycling ratios and spatial scales are associated with vegetation type, position on continent relative to prevailing winds, proximity to mountains, etc. that the authors need to support their 50km definition more strongly.
We opted for our definition of local moisture recycling as we needed a systematic definition of local moisture recycling as we focus on moisture recycling across the globe. We fully agree that factors such as orography, latitude and vegetation affect local moisture recycling. Therefore, the aim of our study is to contribute to the understanding of the relation between local moisture recycling and these different factors. However, we will better discuss the role of landscape characteristics on local moisture recycling and how this relates to the spatial scale of local moisture recycling.
To summarize, we will expand our introduction and discussion. The former to clarify the novelty of our research and the added value of local moisture recycling, and to better embed our work in current literature. The latter to highlight and discuss some assumptions of our analysis. Furthermore, we will conduct some extra analysis to improve the quality of our manuscript. We will do some additional spearman correlation analyses and we will make a comparison with the results from Link et al. We are grateful for the useful comments you provided to our manuscript. We believe they will help us to improve the quality of our work.
On behalf of all authors,
Jolanda Theeuwen
Citation: https://doi.org/10.5194/egusphere-2022-612-AC2
-
AC2: 'Reply on RC2', Jolanda Theeuwen, 04 Oct 2022
Interactive discussion
Status: closed
-
CC1: 'Comment on egusphere-2022-612', Ruud van der Ent, 24 Aug 2022
Dear authors,
I am happy to see that regional recycling at the global scale gets some renewed attention. However there are a couple of points I wish to bring to your attention:
- what's the reason to call your metric 'local' recycling? Isn't it just regional recycling for grid cells of 0.5 arcdegree x 0.5 arcdegree?
- I find the novelty somewhat overstated. As far as I understand the novelty is simply the fact that you calculate regional recycling on a higher resolution grid than other global studies, but the conceptual calculation is very old (see reference list in van der Ent and Savenije (2011) for example).
- a gridcell of 0.5 arcdegree in let's say Stockholm is two times smaller in area than a gridcell around the equator
- moreover, that same grid cell is 3 times smaller in length in east-west direction, and, therefore the dominant wind direction is rather influential on its value.
- in other words, the regional recycling metric or LMR is scale and shape dependent and as such its values cannot be compared from region to region.
- the search for a relation between LMR and other quantities that do not suffer from the scale and shape dependency (precipitation, evaporation, CAPE, biomes etc. ) is therefore fundamentally skewed.
- In Van der Ent and Savenije (2011) I had a suggested alternative metrics, which actually have local meaning for the recycling process, which are the local length scale of precipitation recycling and the local length scale of evaporation recycling. Surely these also rely on a few assumptions, but they do not suffer (or at least to a much more limited extent) from the scale and shape dependency. Please consider this approach or think of a better way to make your metrics scale and shape independent.
Citation: https://doi.org/10.5194/egusphere-2022-612-CC1 -
CC2: 'Reply on CC1', Ruud van der Ent, 24 Aug 2022
Correction, the length of a gridcell in Stockhom (at 59 deg N) is of course also around 2 times smaller than a gridcell a the equator and not 3 times, although further north gridcells that are 3 or more times smaller obviously do exist. Nonetheless the same arguments hold.
Citation: https://doi.org/10.5194/egusphere-2022-612-CC2 -
AC1: 'Reply on CC1', Jolanda Theeuwen, 16 Sep 2022
Dear Dr. van der Ent,
Thank you for your time and the comments you provided on our manuscript: ‘Local moisture recycling across the globe’. Your comments are of great value to improve our manuscript. Below we respond to each point separately to discuss how we will implement them.
what's the reason to call your metric 'local' recycling? Isn't it just regional recycling for grid cells of 0.5 arcdegree x 0.5 arcdegree?
Currently, in literature a similar concept (regional recycling) is used to describe recycling over areas with varying size. Such an area is often related to a peninsula, catchment, etc. Our metric is different from this as we provide information on a more specific and smaller scale namely 0.5 degrees. This is currently the best resolution available for such data with global coverage. We refer to our new metric as the local moisture recycling ratio, first, to highlight the difference with previous studies on regional recycling. Second, this resolution approaches a length scale where local processes have relevant contributions. Feedbacks between the land surface and atmosphere at this spatial scale are relevant for regreening projects. So, we decided to call our metric local recycling as we see potential to use it to study local feedbacks.
I find the novelty somewhat overstated. As far as I understand the novelty is simply the fact that you calculate regional recycling on a higher resolution grid than other global studies, but the conceptual calculation is very old (see reference list in van der Ent and Savenije (2011) for example).
We see there are some similarities to the work done by Van der Ent and Savenije (2011), yet we believe our work is novel. In our paper we assess the impact of the spatial scale within which the moisture recycles by calculating three different types of local moisture recycling. Furthermore, we assess potential drivers of local moisture recycling and by doing so we contribute to the physical understanding of local recycling. However, we will include the work from Van der Ent and Savenije (2011) in our introduction and we will discuss in more detail how our work compares to the work done by Van der Ent and Savenije. Furthermore, we will explain in more detail how our analyses add to the current knowledge, i.e., we can better understand the spatial patterns in local recycling by assessing its drivers. Additionally, the previous work was conducted using output from a different moisture tracking model (i.e., WAM2-layers), which assumes complete mixing of evaporated moisture within two atmospheric layers. For short time scales complete vertical mixing might not be realistic. The UTrack model distributes the evaporated moisture along the vertical moisture profile and therefore, might be more suitable for analyses on a smaller scale. However, we are thankful for this comment which helps us specifying the novelty of our research better.
The following points will be addressed after the last point.
A gridcell of 0.5 arcdegree in let’s say Stockholm is two times smaller in area than a gridcell around the equator
moreover, that same grid cell is 3 times smaller in length in east-west direction, and, therefore the dominant wind direction is rather influential on its value.
In other words, the regional recycling metric or LMR is scale and shape dependent and as such its values cannot be compared from region to region.
The search for a relation between LMR and other quantities that do not suffer from the scale and shape dependency (precipitation, evaporation, CAPE, biomes etc. ) is therefore fundamentally skewed.
In Van der Ent and Savenije (2011) I had a suggested alternative metrics, which actually have local meaning for the recycling process, which are the local length scale of precipitation recycling and the local length scale of evaporation recycling. Surely these also rely on a few assumptions, but they do not suffer (or at least to a much more limited extent) from the scale and shape dependency. Please consider this approach or think of a better way to make your metrics scale and shape independent.
We see that the spatial scale affects moisture recycling ratios, and we believe this is a very useful comment that will help us to improve the quality of our manuscript. To assess this effect for our results, we scaled the local moisture recycling ratio of each grid cell to an area of 50 km x 50 km (see supplement for Figure R1). The relatively large difference between local recycling and scaled local recycling at high latitudes indicates that the local moisture recycling is more uncertain for higher latitudes. For the rest of the globe, we find that the general patterns of local recycling we describe in our manuscript are consistent with and without scaling. The terrestrial surface at lower and mid latitudes are most important for moisture recycling and as the pattern of the scaled and non-scaled local recycling ratio are similar here, we believe the grid cell size causes only a small bias here.
Besides scaling, Van der Ent and Savenije (2011) presented another metric to describe the local moisture recycling, namely, the length scale of evaporation recycling (we will refer to this metric as length scale). They found that this metric scales with their definition of local recycling ratio and has a value typically in the order of 1000 km globally. We did calculate this length scale for our data (see supplement for Figure R2), and its patterns are similar to the patterns we found for local recycling with large values (i.e., small length scales) over tropics and mountainous regions and small values (i.e., large length scales) over desert areas. Similar to the result from Van der Ent and Savenije (2011) we find length scales in the order of 1000 km. Although it is an important metric, we believe that it is more difficult to apply it to, for instance, the impact of land use change on precipitation locally. For this, one needs to determine the amount of rain that recycles locally, and length scale does not quantify this whereas the local moisture recycling ratio does. Yet, to apply local moisture recycling locally, we believe it is important to better understand the local moisture recycling ratio first, and with our study we add to its understanding. In the discussion of our paper, we will discuss the differences between the local moisture recycling ratio and this length scale to allow readers to better assess what metric to use when addressing research questions related to moisture recycling as we do see value in using the length scale for research questions related to non-local effects.
Concerning the comment on the impact of wind direction, we fully agree that the dominant wind direction affects the value of LMR. We can imagine that especially on higher latitudes, where variation between grid cell shape and size is relatively large and the difference between the zonal and meridional length of the grid cell is large, wind direction might have a strong impact on local recycling as you indeed pointed out. We will clarify the impact of dominant wind direction in our discussion to create awareness of this effect amongst the readers.
Finally, you state that the analysis in which we study correlations between the local recycling and different variables is fundamentally skewed due to the scale and shape dependency, which some of the other variables don’t have. We agree that the set-up of our study may result in skewed output, which is intrinsic to our data. Therefore, there is no perfect method to make a comparison between recycling ratios in different regions to study the correlation between the local moisture recycling ratio and other variables. However, to address research questions related to quantifying local hydrology, the local moisture recycling ratio is useful and currently no other metric is available to quantify this globally. A comparison of the recycling ratio among different grid cells is difficult to physically interpret (Van der Ent and Savenije, 2011). Therefore, we need to gain a better understanding of the local moisture recycling ratio. We build this understanding by identifying some of its drivers. However, based on this comment we plan to account for this effect by conducting another analysis in which we classify the data based on latitude and calculate correlation coefficients for the data in these different classes. The grid cell sizes within each class will then be more comparable to minimize the skewness of the analysis. Furthermore, because mainly the higher latitudes are skewed, we excluded Antarctica from our analysis.
To summarize, we will use your comments to address the issue of scaling in more detail in our manuscript. This allows readers to put our results better in perspective. To support this, we will add figures R1 and R2 in the appendix of our manuscript. We would like to thank you for your constructive feedback as it is valuable for improving our manuscript.
On behalf of all authors,
Jolanda Theeuwen
-
CC3: 'Reply on AC1', Ruud van der Ent, 12 Oct 2022
The authors are doing a great job in giving prompt replies rather than waiting until after the public discussion, which illustrates that they're on top of things. I am generally satisfied with their responses, but would like to add a few constructive comments, which are attached in the supplement. I look forward to reading a final version of this manuscript when the formal review process is completed.
-
AC4: 'Reply on CC3', Jolanda Theeuwen, 18 Oct 2022
Dear Dr. Ruud van der Ent,
We are very thankful for your final comments, which are helpful. We will take them into account when we get the opportunity to revise our manuscript.
On behalf of all authors,
Jolanda Theeuwen
Citation: https://doi.org/10.5194/egusphere-2022-612-AC4
-
AC4: 'Reply on CC3', Jolanda Theeuwen, 18 Oct 2022
-
CC3: 'Reply on AC1', Ruud van der Ent, 12 Oct 2022
-
CC2: 'Reply on CC1', Ruud van der Ent, 24 Aug 2022
-
RC1: 'Comment on egusphere-2022-612', Anonymous Referee #1, 15 Sep 2022
Summary
Theeuwen et al. illustrate the local moisture recycling ratio (LMR) globally, i.e. the fraction of evaporation that rains out over the same region or its neighborhood. Using simulations from UTrack driven with ERA5 reanalysis from 2008-2017, they compare the LMR over different neighborhood sizes and illustrate its seasonality. To unravel the drivers of LMR, they calculate the Spearman correlation of the defined LMR to other variables, such as orography, latitude, convective available potential energy, and so on. The discussion centers around the use of these LMR estimates to guide land- and water management practices.
RecommendationThe manuscript is generally well written, but it is, in my opinion, missing novelty and/or a fresh perspective on moisture recycling indices that aim to guide land- and management strategies. Thus, the manuscript requires some major revisions and potentially a slightly different direction to make it a novel and interesting contribution. I will elaborate on my major concerns below.
Major points1. Novelty
The manuscript repeatedly claims that local moisture recycling ratios are calculated "for the first time" (l. 9) and that "it is unknown which fraction of moisture recycles within its source location, and how this recycling varies across the globe" (l. 22-23). However, this is not the first study to do exactly this: Van der Ent et al. (2010) and Van der Ent & Savenjie (2011) already featured such local evaporation recycling ratios and calculated them globally. Furthermore, 'evaporationsheds' (see e.g. Van der Ent & Savenije, 2013) contain the exact same information and papers and data sets have been published on this, see e.g. Link et al. (2020).Unfortunately, I also cannot consider the approach or the objective referred to in the discussion novel: the perspective on understanding the potential influence of land cover, and land- and water management practices via moisture recycling is not new either. Keys et al. (2016), for example, describe this in the context of 'ecosystem services' or 'water security' (Keys et al. 2020) - to name just a few examples. And this is also the subject of all 'green water' studies (e.g., te Wierik et al., 2021; te Wierik et al. 2020).
2. Moisture recycling drivers
I do, however, like the idea of looking at the drivers of moisture recycling; but the current analysis of the drivers is rather simple. In particular, I am a bit hesitant about the variables used to unravel the drivers of LMR, and the methodology used to do so. First of all, while I understand that there is a latitudinal dependence of moisture recycling, I wonder if 'latitude' is the real driver here. Shouldn't it rather be wind, incoming solar radiation and maybe even the underlying area of a grid cell (that differs with latitude)? Similarly, is it fair to use 'evaporation' and 'precipitation' as drivers of LMR? Isn't LMR defined based on these two fluxes? Of course, there is a dependency on both fluxes then... Second, calculating (globally averaged?) Spearman correlations to unravel drivers of LMR is a cheap way of doing this. LMR and any variable in Tab. 1 may be correlated through a third variable that represents the 'true' driver. Or in other words: a correlation does not imply causality.
3. Issues of scale
The definition of what is considered 'local' is rather random. The authors claim that the LMR is based on approx. 50km around the source; however, they also illustrate different definitions of this scale parameter, i.e. 1 grid cell, 9 grid cells and 25 grid cells. The argument for chosing 9 grid cells is rather vague: "To keep the spatial scale as small as possible but to still have a spatial pattern that we can explain physically" (l. 88-89). Could the authors explain why other patterns cannot be explained physically? Is there some lower limit to what the forcing and/or the model can represent? If so, could this limit be determined in a reasonable manner?Some suggestions
To make this a novel and interesting contribution in the field of moisture recycling, a bit more effort may be needed. The authors could, for example, compare their evaporation recycling ratios to the ones from Link et al. - I assume that much more could be learned from the difference of these data sets. Alternatively, the 'true' drivers of moisture recycling could be assessed, using a more sophisticated method to do so. Or the issue of scale and what can be considered local, given the spatio-temporal resolution of the forcing, could be put into focus... these are, however, just some suggestions that I could envision and that would make this paper novel and interesting to me. The authors do not need to follow those.
Minor points- l. 52-53: "Parcels are tracked for up to 30 days or up to the point at which only 1% of their original moisture is still present. " - can this be longer than 30 days?
- Eq. 1-3: this refers to different areas across the globe; where do the 50km from the abstract come in?
- Uncertainty of UTrack is not assessed at all; at least assumptions in the model should be summarised in the Methods section as well.
- Fig. 1: it should probably be "grid cell" and not "grids" in the subtitles
- l. 85-85: "These results seem to indicate that the tracking method we use is not sufficient to define recycling within one grid cell."; maybe it's not the tracking method but the (temporal) resolution of the forcing that is used, or the number of parcels tracked?
- l. 86-87: "Finally, scaling recycling to the number of grid cells, we find r9 and r25 do not relate linearly." Could you elaborate how you scaled this? P is not uniformely distributed across the 9 or 25 grid cells considered here, so I would not expect that there is a linear relationship?
- A suggestion: a uniform color scheme for Figs. 1-2 would be helpful
- l. 111-114: "Both convective and large-scale precipitation correlate with LMR (Table 1), however neither the fraction of convective precipitation nor the fraction of large-scale precipitation correlates with LMR (Table 1). Furthermore, evaporation correlates positively with LMR (Table 1, Fig. 3) indicating that the strong relation between P and LMR is not the only factor that causes a correlation between wetness and LMR." - does it make sense to correlate LMR with P? And as LMR is based on E and P, it needs to be correlated to E as well, right? E could also correlate with LMR because of P... there are so many dependencies here that it is difficult to unravel the real drivers.
- The relation between LMR and convection is not surprising; however, what would be novel was if large-scale and convective precipitation were tracked separately...
- l. 146ff: Are the correlations, especially with convective precipitation and large scale precipitation, subject to spatial and temporal scales?
- l. 155-174: discussion on biomes and deforestation a bit misplaced; not motivated in the introduction at all
- l. 179-181: well-mixed assumption is often hidden in many tracking studies; as far as I understand this is also the case for UTrack - and a recent study illustrated the impact using another Lagrangian model (Keune et al., 2022)
- l. 185ff: this should really be described in the methods, in my opinion
- l. 207: relation to agricultural water management remains unclear to me
- l. 234f: while I understand that you aim to use the LMR as a proxy for regions, in which land and water management may help foster moisture recycling, I don't think this scales at all. To assess the potential of the LMR as a proxy, it would be useful to know if, e.g. an increase in local E by say 10% also leads to an increase of local P by 10%. As you discuss correctly: there are many more factors that play a role here - not just the average recycling ratio; and I am missing an attempt to look at the 'true' drivers of LMR or at least an analysis that moves towards a better suited proxy to estimate the benefit/loss of water due to land- and water management practices in the conext of moisture recycling...
References- Keune, J., Schumacher, D. L., and Miralles, D. G. (2022), A unified framework to estimate the origins of atmospheric moisture and heat using Lagrangian models. Geosci. Model Dev., 15(5), 1875-1898.
- Keys, P. W., Porkka, M., Wang-Erlandsson, L., Fetzer, I., Gleeson, T., and Gordon, L. J. (2019), Invisible water security: Moisture recycling and water resilience, Water Security, 8, 100046.
- Keys, P. W., Wang-Erlandsson, L., and Gordon, L. J. (2016), Revealing invisible water: moisture recycling as an ecosystem service, PloS one, 11(3), e0151993.
- Link, A., van der Ent, R., Berger, M., Eisner, S., and Finkbeiner, M. (2020), The fate of land evaporation – a global dataset, Earth Syst. Sci. Data, 12, 1897–1912, doi:10.5194/essd-12-1897-2020.
- te Wierik, S. A., Cammeraat, E. L. H., Gupta, J., & Artzy-Randrup, Y. A. (2021), Reviewing the impact of land use and land-use change on moisture recycling and precipitation patterns, Water Resour. Res., 57, e2020WR029234. doi:10.1029/2020WR029234.
- te Wierik, S. A., Gupta, J., Cammeraat E. L. H., Artzy-Randrup, Y. A., (2020), The need for green and atmospheric water governance, WIREs Water, 7:e1406, doi:10.1002/wat2.1406.
- van der Ent, R. J., Savenije, H. H. G., Schaefli, B., and Steele-Dunne, S. C. (2010), Origin and fate of atmospheric moisture over continents, Water Resour. Res., 46, W09525, doi:10.1029/2010WR009127
- van der Ent, R. J., and Savenije, H. H. G. (2013), Oceanic sources of continental precipitation and the correlation with sea surface temperature, Water Resour. Res., 49, 3993– 4004, doi:10.1002/wrcr.20296.
- van der Ent, R. J. and Savenije, H. H. G. (2011), Length and time scales of atmospheric moisture recycling, Atmos. Chem. Phys., 11, 1853–1863, doi:10.5194/acp-11-1853-2011.
Citation: https://doi.org/10.5194/egusphere-2022-612-RC1 -
AC3: 'Reply on RC1', Jolanda Theeuwen, 06 Oct 2022
Dear reviewer,
Thank you for your time and for reviewing our manuscript: ‘Local moisture recycling across the globe’. We appreciate your feedback, which is very helpful to improve our manuscript. Below we respond to each major point of feedback separately to discuss how we will implement them. The minor comments will be addressed when we get the opportunity to revise our manuscript.
- Novelty
The manuscript repeatedly claims that local moisture recycling ratios are calculated "for the first time" (l. 9) and that "it is unknown which fraction of moisture recycles within its source location, and how this recycling varies across the globe" (l. 22-23). However, this is not the first study to do exactly this: Van der Ent et al. (2010) and Van der Ent & Savenjie (2011) already featured such local evaporation recycling ratios and calculated them globally. Furthermore, 'evaporationsheds' (see e.g. Van der Ent & Savenije, 2013) contain the exact same information and papers and data sets have been published on this, see e.g. Link et al. (2020).
Unfortunately, I also cannot consider the approach, or the objective referred to in the discussion novel: the perspective on understanding the potential influence of land cover, and land- and water management practices via moisture recycling is not new either. Keys et al. (2016), for example, describe this in the context of 'ecosystem services' or 'water security' (Keys et al. 2020) - to name just a few examples. And this is also the subject of all 'green water' studies (e.g., te Wierik et al., 2021; te Wierik et al. 2020).
We thank the reviewer for explaining why they believe our manuscript is not as novel as stated in the manuscript. Considering that the comment posted by Dr. Ruud van der Ent, the review posted by Dr. Patrick Keys and your review all include this point of feedback, we see the importance of improving on this point. Therefore, we will better acknowledge all relevant previous studies, for example the studies mentioned by the reviewer. Following this we will be better capable of highlighting the differences between our work and the previous work as we do believe there are important new steps being made in our paper. Namely, first, we study the effect of its spatial scale on local moisture recycling. of our definition of local, i.e., the area within which the moisture recycles, and second, we assess potential drivers of local moisture recycling. Besides, the datasets that were used to calculate local recycling in previous studies differ from the dataset used in our study. To obtain these datasets different models were used and the forcing data of these models has a different spatial resolution. We will mention this in our introduction and come back to it in our discussion in more detail by comparing it to the earlier work done by Van der Ent and Savenije (2011). Furthermore, we will highlight how our work compares but also deviates to the work done related to ‘ecosystem services’, ‘water security’ and ‘green water’ studies, to which the reviewer refers Those studies have a focus on source-sink relations in which the sink, apart from the source region, includes also remote locations. In contrast, our work aims to quantify and better understand local recycling. In contrast our work focusses on to quantify and better understand local recycling. The main reason is that previous research on atmospheric moisture connections mainly focusses on non-local water management even though research shows regreening can cause local drying. This suggests the relevance of studying the impact of land cover changes on the local water cycle. Local moisture recycling can help us here. Even though Van der Ent and Savenije (2011) calculated a similar type of recycling the link with preventing local drying has not been made yet, which we believe could be highly valuable. In addition, the spatial scale of 0.5 degrees allows better to study local impacts than the scale of 1.5 degrees. We are thankful for this comment as it helps us to specify the novelty of our manuscript better. We will do this by better acknowledging relevant previous studies.
- Moisture recycling drivers
I do, however, like the idea of looking at the drivers of moisture recycling; but the current analysis of the drivers is rather simple. In particular, I am a bit hesitant about the variables used to unravel the drivers of LMR, and the methodology used to do so. First of all, while I understand that there is a latitudinal dependence of moisture recycling, I wonder if 'latitude' is the real driver here. Shouldn't it rather be wind, incoming solar radiation and maybe even the underlying area of a grid cell (that differs with latitude)? Similarly, is it fair to use 'evaporation' and 'precipitation' as drivers of LMR? Isn't LMR defined based on these two fluxes? Of course, there is a dependency on both fluxes then... Second, calculating (globally averaged?) Spearman correlations to unravel drivers of LMR is a cheap way of doing this. LMR and any variable in Tab. 1 may be correlated through a third variable that represents the 'true' driver. Or in other words: a correlation does not imply causality.
We are happy that the reviewer likes the aim to understand the drivers of local moisture recycling but also understand that the reviewer likes to see more regional (latitudinal) tests that can potentially provide more understanding. In our study we aimed to identify non-linear relations between two variables as most processes cannot be properly described using linear relations and therefore, we used Spearman rank correlations. However, of course we agree that correlation does not imply causality and we will clarify this in the discussion of our manuscript. Moreover, we will use literature to discuss our findings from a mechanistic point of view (see also reviewer Keys). Furthermore, we agree that latitude is not the actual driver of moisture recycling, but that other variables, that correlate with latitude, drive local recycling. As such, we included latitude as a proxy for a combination of processes that have a strong latitudinal pattern. This was not properly described in our manuscript and therefore, we will clarify this in our revision.
Further we will also add more drivers to study the correlation between local moisture recycling and other potential drivers of local moisture recycling, such as solar radiation as suggested by the reviewer. We will not move to multiple regression models, as also clearly indicated by the reviewer it is true that many drivers themselves are correlated to each other. Therefore, we will keep the Spearman rank correlation test per driver. In addition, we plan to split the data in classes based on latitude to account and can then better understand how drivers can change per latitude class which can help to understand the causality. We believe adding more variables to our analysis will improve the understanding of the drivers of local moisture recycling and we are thankful for this comment.
- Issues of scale
The definition of what is considered 'local' is rather random. The authors claim that the LMR is based on approx. 50km around the source; however, they also illustrate different definitions of this scale parameter, i.e. 1 grid cell, 9 grid cells and 25 grid cells. The argument for chosing 9 grid cells is rather vague: "To keep the spatial scale as small as possible but to still have a spatial pattern that we can explain physically" (l. 88-89). Could the authors explain why other patterns cannot be explained physically? Is there some lower limit to what the forcing and/or the model can represent? If so, could this limit be determined in a reasonable manner?
We thank the reviewer for pointing out the unclarity concerning the spatial scale of local moisture recycling. We agree that the definition is partly arbitrary. We will mention this in our manuscript. Concerning the decision to use recycling over 9 grid cells, local recycling within one grid cell results in exceptionally low values over mountain peaks, yet not over all elevated terrain and relatively high values over the ocean. This pattern is inconsistent with the result found for recycling within 9 and 25 grid cells. The patterns for recycling over 9 and 25 grid cells can be explained as high values over mountains can result from convection as a result of orographic lift and relatively low values over ocean can be explained by the large atmospheric moisture transport due to strong winds. Possibly a numerical process is in place for the recycling within one grid cell, causing the pattern to be different from recycling over 9 and 25 grid cells. As the pattern of the latter two definitions are similar and agree with our understanding we decided to define local moisture recycling as the recycling over evaporated moisture within its source grid cell and its 8 surrounding grid cells. We will clarify this part in our results section. Furthermore, we will rephrase the sentence in which we state we cannot “physically” explain the pattern of recycling within one grid cell. We will omit the word physically and state we cannot fully explain the pattern of recycling within one grid cell. We believe these adjustments will clarify our decision and we thank the reviewer for pointing this out.
Some suggestions
To make this a novel and interesting contribution in the field of moisture recycling, a bit more effort may be needed. The authors could, for example, compare their evaporation recycling ratios to the ones from Link et al. - I assume that much more could be learned from the difference of these data sets. Alternatively, the 'true' drivers of moisture recycling could be assessed, using a more sophisticated method to do so. Or the issue of scale and what can be considered local, given the spatio-temporal resolution of the forcing, could be put into focus... these are, however, just some suggestions that I could envision and that would make this paper novel and interesting to me. The authors do not need to follow those.
We agree with these suggestions. As described before we will conduct more analyses by studying the relation between local moisture recycling and more potential drivers (e.g., solar radiation and other aspects of the energy balance). In addition, we will classify our data based and for each latitudinal class and conduct a spearman rank correlation analysis separately. Furthermore, we will compare our results to the dataset by Link et al. (2020). This means we need to upscale our findings to the 1.5 degrees resolution to match the dataset from Link et al (2020) to the dataset by Tuinenburg et al. (2020). This will give some insight into potential differences due to different use of models and simulated time periods, independent to the resolution used.
We would like to thank the reviewer for their constructive feedback as it is valuable for improving our manuscript. As mentioned before, in a future response will address the minor points of the reviewer.
On behalf of all authors,
Jolanda Theeuwen
References:
Link, A., van der Ent, R., Berger, M., Eisner, S., and Finkbeiner, M.: The fate of land evaporation – a global dataset, Earth Syst. Sci. Data, 12, 1897–1912, https://doi.org/10.5194/essd-12-1897-2020, 2020.
Van der Ent, R. J., & Savenije, H. H. G. (2011). Length and time scales of atmospheric moisture recycling. Atmospheric Chemistry and Physics, 11(5), 1853–1863. https://doi.org/10.5194/acp-11-1853-2011
Van der Ent, R. J., Savenije, H. H. G., Schaefli, B., & Steele-Dunne, S. C. (2010). Origin and fate of atmospheric moisture over continents. Water Resources Research, 46(9), W09525. https://doi.org/10.1029/2010WR009127
Keys, P. W., Wang-Erlandsson, L., and Gordon, L. J. (2016), Revealing invisible water: moisture recycling as an ecosystem service, PloS one, 11(3), e0151993.
te Wierik, S. A., Cammeraat, E. L. H., Gupta, J., & Artzy-Randrup, Y. A. (2021), Reviewing the impact of land use and land-use change on moisture recycling and precipitation patterns, Water Resour. Res., 57, e2020WR029234. doi:10.1029/2020WR029234.
te Wierik, S. A., Gupta, J., Cammeraat E. L. H., Artzy-Randrup, Y. A., (2020), The need for green and atmospheric water governance, WIREs Water, 7:e1406, doi:10.1002/wat2.1406.
Tuinenburg, Obbe A., Theeuwen, J. J. E., & Staal, A. (2020). High-resolution global atmospheric moisture connections from evaporation to precipitation. Earth System Science Data S, 12(4), 3177–3188. https://doi.org/10.5194/essd-12-3177-2020
Citation: https://doi.org/10.5194/egusphere-2022-612-AC3
-
RC2: 'Comment on egusphere-2022-612', Patrick Keys, 16 Sep 2022
SUMMARY
The authors of "Local moisture recycling across the globe" explore the concept of local moisture recycling (LMR), and its distribution and characteristics globally. The authors aim to understand the drivers of LMR and explore the heterogeneity of the phenomena across latitudes, biomes, and elevations. In general, I find the work interesting and the analyses mostly sound. I think that once the authors can make some changes — which I think would fall into the minor category, but I suppose could be major — this ought to be published.Here are several general comments followed by more line specific comments afterward.
GENERAL COMMENTS
1. "Drivers": I had some trouble with the word “drivers” being used to describe the role of e.g., CAPE, in LMR. While it's possible that some of the correlated variables could be causally related to LMR, there was no causal analysis completed (or a mechanistic explanation) as far as I could tell. Given that, it seems perfectly reasonable to say correlated phenomena (or similar) with regard to the variables that seemed to have convincing correlations with LMR, e.g., wetness or elevation.2. LMR Definition: This is purely a suggestion, but I would strongly encourage the authors to frame the LMR idea conceptually first (which I think would force the authors to argue more clearly for the novelty of the idea), and then provide the specific way that they define it in the article (e.g., "LMR is _________, which we define here as ______"). The reason being that LMR could be a conceptually useful idea on its own, but others may make entirely different quantifications, or indeed develop a more robust geophysical definition later on. By distinguishing your conceptual contribution from the physical definition, you may give the concept more scholarly and applied longevity. Again, just a suggestion.
3. Expand the Introduction: Given the intended scope of this work, the introduction should be expanded to include discussion of past moisture recycling analyses that relate to the LMR idea. There are quite a lot of moisture recycling studies that examine: local recycling (though perhaps not at the global scale), the scale dependence of moisture recycling, and the role of different types of vegetation in the scale dependence of moisture recycling (e.g., van der Ent et al., 2014).
4. Discussion of robustness of a 10-yr climatology: The current draft of the paper does not suitably discuss the appropriateness of the time scale associated with the Utrack dataset in the context of climate variation. Given that the authors are making claims about 'average rates', correlating with phenomena to determine their relationship (potentially causative) of LMR, etc., it is necessary to provide both a justification and a discussion (including of limitations) vis a vis a 10-year climatology. Given that there are inter-annual (e.g., ENSO, Indian Ocean Dipole) and decadal-scale (e.g., Pacific Decadal Oscillation), modes of climate variability that could be systematically affecting some of these results, it is necessary to explain the role of a 10-yr only analysis. Saying that the data are only available for 10 years is not really sufficient. Acknowledging this temporal limitation is critical also for making sure readers can appropriately interpret the results, which could change with a longer time series. Again — I recognize the authors have mentioned the topic of the length of record albeit briefly, but a more in-depth discussion is needed.
5. A 50-km definition: The authors select 50 km as the spatial scale of their LMR definition. A bit more discussion is needed to explain the logic and rationale of why such a blanket definition across the terrestrial surface is appropriate, and not an orographically-, latitudinally-, or biome-dependent definition. I’m not suggesting to change away from 50 km, but enough other research has found that moisture recycling ratios and spatial scales are associated with vegetation type, position on continent relative to prevailing winds, proximity to mountains, etc. that the authors need to support their 50km definition more strongly.
SPECIFIC COMMENTS
L10 Consider saying “defined here as…”L11 You could consider using the phrasing ”a 10-year climatology…”
L21 See General Comment about Expanding the Introduction.
L23 The sentence beginning with “However, it is unknown…” seems like it might not truly reflect the state of knowledge about moisture recycling globally, and overclaim the knowledge gap. You could consider situating this more concretely in what is understood (e.g., from the perspective of vegetation type having different length scaling associated with moisture recycling, etc.)
L38 The authors propose to explore “spatial-temporal variation across the globe” but a 10 year climatology is not enough to support the “temporal variation” component.
L67 What is the origin of these correlated variables? I could not see whether they came from ERA5 or somewhere else. Likewise, given the range in geophysical process scale of these variables, it would be good to comment here (or in the discussion of limitations) on the spatial scale of some of these, and whether some are more or less appropriate at the scales of analysis used in the study.
L67 See General Comment about the word “Drivers.”
L78 It is worth noting (either here, or in the Discussion) that the specific continental configuration (and mountain ranges) relative to prevailing winds and the associated biome composition matters quite a lot for determining the distribution of moisture recycling ratios. I know that the authors know this already (just looking at the author names and their past publications) so it is a noticeable gap in the logic.
Fig3 This is an interesting figure, but I think that the bottom left panel provides the most insight. I would recommend pulling this figure out on its own, so that it can be seen in a larger capacity, given the density of dots and information presented.
L135 Again, drivers is probably not quite right here, since there is no sufficiently causal mechanistic explanation of the link between e.g., CAPE and LMR.
L141 This seems like a perfect opportunity to situate the findings in the sweep of convective storm literature. I suspect that a convective storm meteorologist might not find the statement very surprising “our results suggest a positive relation between convection and LMR.” That doesn’t mean it shouldn’t be said — just that some references aligning this statement with the corresponding literature seem prudent.
L149 If I’m looking at the right Miyamoto 2013 reference, the argument appears to be that the number of convective features increases dramatically with resolution. That being said, the statement that “convection is a local scale process (i.e., spatial scale of 100 km)” might want to be adjusted.
L145 The entire paragraph needs better referencing, since the authors are making numerous claims regarding convective storms, and how the LMR analysis relates to that field. Greater referencing would also give me (the reader) more confidence that these claims are supported by the broader field.
L155 The global biome discussion is very interesting, but in its current presentation it both (a) reads as results, and (b) needs a supporting figure in the main text.
L168 These findings are very interesting, and are well discussed. I would use the density of supporting references and citations here as an example of what is necessary in the “convection” section above.
L203 The authors rely heavily (though not entirely) on Salmon et al 2011. Given the range of claims being discussed here, it seems prudent to include a few more references (than relying on Salmon et al. three times in the same paragraph.
L221 I encourage the authors to cite the work by Kirsten Findell in this paragraph (see Refs below), who provides a global analysis which blends empirical analysis and theory to explore how continental moisture recycling may change over the coming century.
L235 The sentence should be restructured for clarity.
L242 It might be interesting to be able to state the standard deviation associated with the 1.6% number that is quoted throughout the paper.
REFERENCES
Findell, K.L. et al. (2019) ‘Rising Temperatures Increase Importance of Oceanic Evaporation as a Source for Continental Precipitation’, Journal of Climate, 32(22), pp. 7713–7726.
van der Ent, R.J. et al. (2014) ‘Contrasting roles of interception and transpiration in the hydrological cycle – Part 2: Moisture recycling’, Earth System Dynamics, 5(2), pp. 471–489.
Citation: https://doi.org/10.5194/egusphere-2022-612-RC2 -
AC2: 'Reply on RC2', Jolanda Theeuwen, 04 Oct 2022
Dear Dr. Patrick Keys,
Thank you for taking the time to review our manuscript ‘Local moisture recycling across the globe’. We very much appreciate your feedback as we believe they will help us to improve our manuscript. Below we will shortly respond to each of the general comments you wrote in your review. We will address the specific comments and the general comments in more detail when we get the opportunity to revise our manuscript.
- "Drivers": I had some trouble with the word “drivers” being used to describe the role of e.g., CAPE, in LMR. While it's possible that some of the correlated variables could be causally related to LMR, there was no causal analysis completed (or a mechanistic explanation) as far as I could tell. Given that, it seems perfectly reasonable to say correlated phenomena (or similar) with regard to the variables that seemed to have convincing correlations with LMR, e.g., wetness or elevation.
We agree on the point that a correlation between LMR and any other variable does not imply causality in our study and therefore, we understand that “driver” might not be the best word to use in our manuscript. We thank you for this contribution; we will use different terminology throughout our manuscript and think that the term ‘factors’ is more appropriate. Furthermore, we will highlight in our discussion or methods section that a correlation does not imply causality in our study.
In addition we will better implement our hypothesis into our manuscript. We will explain what processes we expect LMR to be part of using previous literature. This will support our decision what variables we included in our study. For each variable in our study, we can also highlight whether we expect a direct relation between this variable and local moisture recycling or whether this variable is a proxy.
- LMR Definition: This is purely a suggestion, but I would strongly encourage the authors to frame the LMR idea conceptually first (which I think would force the authors to argue more clearly for the novelty of the idea), and then provide the specific way that they define it in the article (e.g., “LMR is _________, which we define here as ______”). The reason being that LMR could be a conceptually useful idea on its own, but others may make entirely different quantifications, or indeed develop a more robust geophysical definition later on. By distinguishing your conceptual contribution from the physical definition, you may give the concept more scholarly and applied longevity. Again, just a suggestion.
We very much appreciate this suggestion from the reviewer. We believe by framing LMR conceptually our proposal to study local hydrological impacts of land cover changes becomes clearer. This allows future studies to explore this novel concept of local moisture recycling using different definitions. We see how this indeed helps us to better highlight the novelty of our approach and we agree that by framing it as you suggest we better clarify the importance of moisture recycling locally regarding land cover change. Many thanks for this suggestion we will implement it in the introduction.
- Expand the Introduction: Given the intended scope of this work, the introduction should be expanded to include discussion of past moisture recycling analyses that relate to the LMR idea. There are quite a lot of moisture recycling studies that examine: local recycling (though perhaps not at the global scale), the scale dependence of moisture recycling, and the role of different types of vegetation in the scale dependence of moisture recycling (e.g., van der Ent et al., 2014).
To better highlight the novelty and aim we will expand the introduction and include more literature on studies related to moisture recycling. We will discuss the length scale of evaporation recycling in our introduction. We also will discuss its spatial patterns and explain its physical meaning. We will use this to clarify that the length scale is different from the local moisture recycling ratio and clearly show the added value of the local moisture recycling to the scientific community is. In our response to the community comment from dr. Ruud van der Ent we already did describe in more detail how we will implement the length scale in the introduction and discussion of our manuscript. Thank you for providing these examples.
- Discussion of robustness of a 10-yr climatology: The current draft of the paper does not suitably discuss the appropriateness of the time scale associated with the Utrack dataset in the context of climate variation. Given that the authors are making claims about ‘average rates’, correlating with phenomena to determine their relationship (potentially causative) of LMR, etc., it is necessary to provide both a justification and a discussion (including of limitations) vis a vis a 10-year climatology. Given that there are inter-annual (e.g., ENSO, Indian Ocean Dipole) and decadal-scale (e.g., Pacific Decadal Oscillation), modes of climate variability that could be systematically affecting some of these results, it is necessary to explain the role of a 10-yr only analysis. Saying that the data are only available for 10 years is not really sufficient. Acknowledging this temporal limitation is critical also for making sure readers can appropriately interpret the results, which could change with a longer time series. Again — I recognize the authors have mentioned the topic of the length of record albeit briefly, but a more in-depth discussion is needed.
We agree that the 10-year averaged data is not sufficient for a robust climatological analysis and will highlight this better in our manuscript. Our analysis, and the 10 years involved, are used for intra-variability (seasonlity) within the year. However, trends in multi-year climate variation could affect our results. To discuss this point, we will compare our results with the results from Link et al. (2020), as that study has a longer time period. To enable a comparison, we will calculate local moisture recycling ratios at a spatial scale of 1.5 degrees from both the dataset by Link et al. (2020) and the dataset by Tuinenburg et al. (2020).
- A 50-km definition: The authors select 50 km as the spatial scale of their LMR definition. A bit more discussion is needed to explain the logic and rationale of why such a blanket definition across the terrestrial surface is appropriate, and not an orographically-, latitudinally-, or biome-dependent definition. I’m not suggesting to change away from 50 km, but enough other research has found that moisture recycling ratios and spatial scales are associated with vegetation type, position on continent relative to prevailing winds, proximity to mountains, etc. that the authors need to support their 50km definition more strongly.
We opted for our definition of local moisture recycling as we needed a systematic definition of local moisture recycling as we focus on moisture recycling across the globe. We fully agree that factors such as orography, latitude and vegetation affect local moisture recycling. Therefore, the aim of our study is to contribute to the understanding of the relation between local moisture recycling and these different factors. However, we will better discuss the role of landscape characteristics on local moisture recycling and how this relates to the spatial scale of local moisture recycling.
To summarize, we will expand our introduction and discussion. The former to clarify the novelty of our research and the added value of local moisture recycling, and to better embed our work in current literature. The latter to highlight and discuss some assumptions of our analysis. Furthermore, we will conduct some extra analysis to improve the quality of our manuscript. We will do some additional spearman correlation analyses and we will make a comparison with the results from Link et al. We are grateful for the useful comments you provided to our manuscript. We believe they will help us to improve the quality of our work.
On behalf of all authors,
Jolanda Theeuwen
Citation: https://doi.org/10.5194/egusphere-2022-612-AC2
-
AC2: 'Reply on RC2', Jolanda Theeuwen, 04 Oct 2022
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 824 | 296 | 30 | 1,150 | 7 | 9 |
- HTML: 824
- PDF: 296
- XML: 30
- Total: 1,150
- BibTeX: 7
- EndNote: 9
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 291 | 26 |
| Netherlands | 2 | 265 | 24 |
| Germany | 3 | 85 | 7 |
| China | 4 | 70 | 6 |
| France | 5 | 36 | 3 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 291
Cited
1 citations as recorded by crossref.
Jolanda Theeuwen
Arie Staal
Obbe Tuinenburg
Bert Hamelers
Stefan Dekker
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(3038 KB) - Metadata XML