Robust global detection of forced changes in mean and extreme precipitation despite observational disagreement on the magnitude of change

de Vries, Iris Elisabeth; Sippel, Sebastian; Pendergrass, Angeline Greene; Knutti, Reto

doi:https://doi.org/10.5194/egusphere-2022-568

Preprints

https://doi.org/10.5194/egusphere-2022-568

Preprints

13 Jul 2022

| 13 Jul 2022

Robust global detection of forced changes in mean and extreme precipitation despite observational disagreement on the magnitude of change

Iris Elisabeth de Vries, Sebastian Sippel, Angeline Greene Pendergrass, and Reto Knutti

Abstract. Detection and attribution (D&A) of forced precipitation change is challenging due to internal variability and limited spatial and temporal coverage of observational records. These factors result in a low signal-to-noise ratio of potential regional and even global trends. Here, we use a statistical method – ridge regression – to create physically interpretable fingerprints for detection of forced changes in mean and extreme precipitation with a high signal-to-noise ratio. The fingerprints are constructed using CMIP6 multi-model output masked to match coverage of three gridded precipitation observational datasets – GHCNDEX, HadEX3, and GPCC –, and are then applied to these observational datasets to assess the degree of forced change detectable in the real-world climate.

We show that the signature of forced change is detected in all three observational datasets for global metrics of mean and extreme precipitation. Forced changes are still detectable from changes in the spatial patterns of precipitation even if the global mean trend is removed from the data. This shows detection of forced change in mean and extreme precipitation beyond a global mean trend, and increases confidence in the detection method's power, as well as in climate models' ability to capture the relevant processes that contribute to large-scale patterns of change.

We also find, however, that detectability depends on the observational dataset used. Not only coverage differences but also observational uncertainty contribute to dataset disagreement, exemplified by times of emergence of forced change from internal variability ranging from 1998 to 2004 among datasets. Furthermore, different choices for the period over which the forced trend is computed result in different levels of agreement between observations and model projections. These sensitivities may explain apparent contradictions in recent studies on whether models under- or overestimate the observed forced increase in mean and extreme precipitation. Lastly, the detection fingerprints are found to rely primarily on the signal in the extratropical Northern Hemisphere, which is at least partly due to observational coverage, but potentially also due to the presence of a more robust signal in the Northern Hemisphere in general.

Received: 29 Jun 2022 – Discussion started: 13 Jul 2022

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.

Download & links

Preprint (PDF, 3433 KB)

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 (3433 KB)

Supplement (9191 KB)

Download & links

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Journal article(s) based on this preprint

26 Jan 2023

| Highlight paper

Robust global detection of forced changes in mean and extreme precipitation despite observational disagreement on the magnitude of change

Iris Elisabeth de Vries, Sebastian Sippel, Angeline Greene Pendergrass, and Reto Knutti

Earth Syst. Dynam., 14, 81–100, https://doi.org/10.5194/esd-14-81-2023,https://doi.org/10.5194/esd-14-81-2023, 2023

Short summary Chief editor

Iris Elisabeth de Vries, Sebastian Sippel, Angeline Greene Pendergrass, and Reto Knutti

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2022-568', Anonymous Referee #1, 02 Sep 2022

This paper proposes a ridge regression approach to the detection and attribution of externally forced changes in mean and extreme precipitation. This is an interesting idea that certainly merits exploration, but before devoting a lot of time to understanding the details of the paper and the results that are obtained, I think it is necessary for the authors to better explain their method and to situate it within the pantheon of methods that are already available for detection and attribution.

Ridge regression is a technique that “regularizes” regression problems, such as that described in equation (1) of the paper, in which the predictor variables contained in matrix X are multi-colinear. In the generalized least squares formulation of the regression used in detection and attribution this matrix is composed of model simulated estimates of the responses to external forcing in the form of space-time patterns of change. Depending on variable, period considered, domain of interest and how data are processed, the expected space-time patterns of responses to different forcing factors (often called fingerprints) can be strongly correlated, which results in a regression “design matrix” X that may be ill-conditioned. Ridge regression is a technique that can be used to overcome this problem, although I imagine at the cost of introducimg some bias into the estimated signal scaling coefficients β. Note that referring to these coefficients as “fingerprints” seems unusual to me.

The concept of regularization, however, also arises in a second way in the detection and attribution problem. Considering again equation (1), the generalized least squares approach (and also its total least squares extension) requires knowledge of the variance-covariance matrix of the residuals ε, which are regarded as resulting from natural internal climate variability. Thus, the variance-covariance matrix is generally estimated from unforced control simulations, using as many climate-model simulated realisations of ε as possible. Even though many climate-model simulated realizations of ε are now generally available, the estimated variance-covariance matrix may not be of full rank or may remain uncertain. Thus, it is also often regularized, using an approach similar to the regularization used in ridge regression, but applied to the noise term rather than the signal term of equation (1). See Ribes et al (2013a, doi:10.1007/s00382-013-1735-7, and 2013b, doi:10.1007/s00382-013-1736-6). Presumably one would want to regularize both aspects of the problem, and also take signal uncertainty into account as is done in the total least squares approach to the regression problem (see again Ribes et al., 2013a and 2013b, and also Allen and Stott, 2003, doi:10.1007/s00382-003-0313-9).

How the combined model represented by equations (1-3) relates to existing techniques, and now the noise that results from internal variability comes into play and is accounted for in their subsequent application in the paper is not made clear, and I think should be clarified before results can be considered.

Also, I think it is necessary for the authors to discuss whether the proposed methods, which basically use linear statistical models that therefore implicitly assume Gaussian, or near Gaussian errors, are suitable for the data to which they are applied. Indicators of extreme precipitation, such as Rx1day at individual grid boxes, are certainly not Gaussian.

A final general comment is that the relatively heavy of use of acronyms in this paper is not very reader friendly.

Citation: https://doi.org/10.5194/egusphere-2022-568-RC1
- AC1: 'Reply on RC1', Iris de Vries, 08 Sep 2022
  
  The authors thank referee #1 for the important comments. The attached pdf contains our reply.
  
  Citation: https://doi.org/10.5194/egusphere-2022-568-AC1
RC2:
'Comment on egusphere-2022-568', Anonymous Referee #2, 06 Sep 2022

Overall comments:

This study conducts a signal detection analysis for global changes in mean and extreme precipitation using three observational datasets and CMIP6 multi-model outputs. The authors apply a ridge regression (RR) method to construct fingerprints, which helps increase a signal-to-noise ratio of precipitation change patterns. Results show a robust detection of anthropogenic signals in all observations for both mean and extreme precipitation even when removing global mean trends, further supporting the human-induced intensification of global hydrological cycle. I find this paper very well written with sufficient details provided about methods as well as various sensitivity tests and therefore suggest publication after addressing some minor issues.

Major comments:

1. Although method details are provided, it would be useful to explain more clearly what are benefits of the attribution approaches employed, including ridge regression, EOF-based metric for target variable, and GMST-based signal estimation. All of these procedures seem to contribute to increase signal-to-noise ratio but how they do and what step is more important. The authors provide some associated results from sensitivity tests but an overall explanation of their method possibly with a schematic would be helpful for readers to understand the contribution of each step to the final signal detection.

2. An important motivation of considering different periods and datasets is opposing conclusions by previous studies about model overestimation or underestimation of the observed trends. I am wondering if the authors can go further and compare their results with some previous studies. For instance, if studies based on the latter half of 20^th century trends find model underestimation, the authors can assess their model trends for the same/similar periods. Another point here is that the present study uses absolute units of precipitation while most of previous studies considered relative changes or aggregated values. It would be good to discuss possible influences of this difference.

3. The lower detectability in GHCNDEX observations are suggested to be due to the poorer spatial coverage. Regarding this issue, I would suggest using Rx5d. As I understand, Rx5d has larger spatial coverage than Rx1d and comparison with Rx1d-based results may provide a way to support the authors’ interpretation. Another way would be to compare detection results from using a selected model run but with different spatial coverages applied.

Minor comments:

L8: Indicating analysis period or trend period with signal detection would be useful here.

L17-19, L58-64: Better comparisons can be made by applying the same periods as those used in previous studies. See my major comment above.

L20-21: Is this confirmed by repeating detection analysis using NH-extratropics only?

L34: “discrepancies with respect to observations”. Its meaning is unclear.

L69-71: Need to explain what the previous studies have found additionally using these “data-science methods”. Also, what’s the novelty of this study compared with them? Is it detection based on spatial pattern information alone?

L108-109: “Trend biases due to this structural difference … negligible”. But the cited reference considered south-east Australia only?

L201: How to define S when global means are removed?

L212: “CMIP6 ssp245” should be “CMIP6 historical”?

L227: “virtually identical”. adding spatial correlation would help with this.

L314-316: This suggests possible dependence of Rx1d FRE on temperature, resembling global warming slowdown due to PDO influence?

L331-332: “results … hold when the global mean is used as FR target”. Then what are benefits of using EOF-based metric for target variable?

L382-383: “accuracy of the CMIP6 climate models in simulating the processes …”. It’s unclear how the authors get this conclusion. Observation-model agreement in residual variability? More explanation would be useful.

L394-395: “(not shown)”. This looks important and I suggest showing them in the supplement.

L428: “value of RR-based fingerprint construction”. What happens in detection or SNR without applying RR? See my major comment above.

Citation: https://doi.org/10.5194/egusphere-2022-568-RC2
- AC2: 'Reply on RC2', Iris de Vries, 12 Oct 2022
  
  The authors thank referee #2 for the useful comments. Our response is attached as a PDF.
  
  Citation: https://doi.org/10.5194/egusphere-2022-568-AC2

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2022-568', Anonymous Referee #1, 02 Sep 2022

This paper proposes a ridge regression approach to the detection and attribution of externally forced changes in mean and extreme precipitation. This is an interesting idea that certainly merits exploration, but before devoting a lot of time to understanding the details of the paper and the results that are obtained, I think it is necessary for the authors to better explain their method and to situate it within the pantheon of methods that are already available for detection and attribution.

Ridge regression is a technique that “regularizes” regression problems, such as that described in equation (1) of the paper, in which the predictor variables contained in matrix X are multi-colinear. In the generalized least squares formulation of the regression used in detection and attribution this matrix is composed of model simulated estimates of the responses to external forcing in the form of space-time patterns of change. Depending on variable, period considered, domain of interest and how data are processed, the expected space-time patterns of responses to different forcing factors (often called fingerprints) can be strongly correlated, which results in a regression “design matrix” X that may be ill-conditioned. Ridge regression is a technique that can be used to overcome this problem, although I imagine at the cost of introducimg some bias into the estimated signal scaling coefficients β. Note that referring to these coefficients as “fingerprints” seems unusual to me.

The concept of regularization, however, also arises in a second way in the detection and attribution problem. Considering again equation (1), the generalized least squares approach (and also its total least squares extension) requires knowledge of the variance-covariance matrix of the residuals ε, which are regarded as resulting from natural internal climate variability. Thus, the variance-covariance matrix is generally estimated from unforced control simulations, using as many climate-model simulated realisations of ε as possible. Even though many climate-model simulated realizations of ε are now generally available, the estimated variance-covariance matrix may not be of full rank or may remain uncertain. Thus, it is also often regularized, using an approach similar to the regularization used in ridge regression, but applied to the noise term rather than the signal term of equation (1). See Ribes et al (2013a, doi:10.1007/s00382-013-1735-7, and 2013b, doi:10.1007/s00382-013-1736-6). Presumably one would want to regularize both aspects of the problem, and also take signal uncertainty into account as is done in the total least squares approach to the regression problem (see again Ribes et al., 2013a and 2013b, and also Allen and Stott, 2003, doi:10.1007/s00382-003-0313-9).

How the combined model represented by equations (1-3) relates to existing techniques, and now the noise that results from internal variability comes into play and is accounted for in their subsequent application in the paper is not made clear, and I think should be clarified before results can be considered.

Also, I think it is necessary for the authors to discuss whether the proposed methods, which basically use linear statistical models that therefore implicitly assume Gaussian, or near Gaussian errors, are suitable for the data to which they are applied. Indicators of extreme precipitation, such as Rx1day at individual grid boxes, are certainly not Gaussian.

A final general comment is that the relatively heavy of use of acronyms in this paper is not very reader friendly.

Citation: https://doi.org/10.5194/egusphere-2022-568-RC1
- AC1: 'Reply on RC1', Iris de Vries, 08 Sep 2022
  
  The authors thank referee #1 for the important comments. The attached pdf contains our reply.
  
  Citation: https://doi.org/10.5194/egusphere-2022-568-AC1
RC2:
'Comment on egusphere-2022-568', Anonymous Referee #2, 06 Sep 2022

Overall comments:

This study conducts a signal detection analysis for global changes in mean and extreme precipitation using three observational datasets and CMIP6 multi-model outputs. The authors apply a ridge regression (RR) method to construct fingerprints, which helps increase a signal-to-noise ratio of precipitation change patterns. Results show a robust detection of anthropogenic signals in all observations for both mean and extreme precipitation even when removing global mean trends, further supporting the human-induced intensification of global hydrological cycle. I find this paper very well written with sufficient details provided about methods as well as various sensitivity tests and therefore suggest publication after addressing some minor issues.

Major comments:

1. Although method details are provided, it would be useful to explain more clearly what are benefits of the attribution approaches employed, including ridge regression, EOF-based metric for target variable, and GMST-based signal estimation. All of these procedures seem to contribute to increase signal-to-noise ratio but how they do and what step is more important. The authors provide some associated results from sensitivity tests but an overall explanation of their method possibly with a schematic would be helpful for readers to understand the contribution of each step to the final signal detection.

2. An important motivation of considering different periods and datasets is opposing conclusions by previous studies about model overestimation or underestimation of the observed trends. I am wondering if the authors can go further and compare their results with some previous studies. For instance, if studies based on the latter half of 20^th century trends find model underestimation, the authors can assess their model trends for the same/similar periods. Another point here is that the present study uses absolute units of precipitation while most of previous studies considered relative changes or aggregated values. It would be good to discuss possible influences of this difference.

3. The lower detectability in GHCNDEX observations are suggested to be due to the poorer spatial coverage. Regarding this issue, I would suggest using Rx5d. As I understand, Rx5d has larger spatial coverage than Rx1d and comparison with Rx1d-based results may provide a way to support the authors’ interpretation. Another way would be to compare detection results from using a selected model run but with different spatial coverages applied.

Minor comments:

L8: Indicating analysis period or trend period with signal detection would be useful here.

L17-19, L58-64: Better comparisons can be made by applying the same periods as those used in previous studies. See my major comment above.

L20-21: Is this confirmed by repeating detection analysis using NH-extratropics only?

L34: “discrepancies with respect to observations”. Its meaning is unclear.

L69-71: Need to explain what the previous studies have found additionally using these “data-science methods”. Also, what’s the novelty of this study compared with them? Is it detection based on spatial pattern information alone?

L108-109: “Trend biases due to this structural difference … negligible”. But the cited reference considered south-east Australia only?

L201: How to define S when global means are removed?

L212: “CMIP6 ssp245” should be “CMIP6 historical”?

L227: “virtually identical”. adding spatial correlation would help with this.

L314-316: This suggests possible dependence of Rx1d FRE on temperature, resembling global warming slowdown due to PDO influence?

L331-332: “results … hold when the global mean is used as FR target”. Then what are benefits of using EOF-based metric for target variable?

L382-383: “accuracy of the CMIP6 climate models in simulating the processes …”. It’s unclear how the authors get this conclusion. Observation-model agreement in residual variability? More explanation would be useful.

L394-395: “(not shown)”. This looks important and I suggest showing them in the supplement.

L428: “value of RR-based fingerprint construction”. What happens in detection or SNR without applying RR? See my major comment above.

Citation: https://doi.org/10.5194/egusphere-2022-568-RC2
- AC2: 'Reply on RC2', Iris de Vries, 12 Oct 2022
  
  The authors thank referee #2 for the useful comments. Our response is attached as a PDF.
  
  Citation: https://doi.org/10.5194/egusphere-2022-568-AC2

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

ED: Reconsider after major revisions (13 Oct 2022) by Gabriele Messori

AR by Iris de Vries on behalf of the Authors (23 Nov 2022) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (24 Nov 2022) by Gabriele Messori

RR by Anonymous Referee #1 (06 Dec 2022)

Suggestions for revision or reasons for rejection

First, I would like to thank the authors for their detailed responses to my comments and for the improvements that have been made to the paper, which in my view, is much improved, particularly regarding the presentation of key concepts that are needed to understand the paper and how it is situated in the “D&A” literature.

I do have some additional specific comments that might help to further improve an already very good paper, which I hope you find useful.

60-77: How does this paper help to resolve the issues that are raised here concerning the robustness of findings from D&A studies and the representation of precipitation related processes in models? The text that begins with “Here we show …” (line 70) doesn’t really answer that question.

116-119: It seems that you’ve tried to create a very “balanced” climate model dataset (3 runs per model, 450 years of PI-control per model for as many models as possible). A few words on the sampling approach (e.g., what motivates it, what problems are avoided through a balanced approach, and what other problems might arise – see comments below) might be appropriate.

134-135: I think the crucial issue is not so much the locations that grid-cells represent, but rather, that quantities that they represent. The question of what quantity they represent is easier to understand in the case of Rx1day. I think we can all agree that the Rx1day values obtained from models can only be interpreted as annual maxima of daily spatial mean (grid box average) precipitation amounts. In contrast, when a product like HadEX3 is produced from station data, Rx1day is first determined at stations and then those point Rx1day values are spatially interpolated to obtain a grid box value. This is a different number from the annual maximum of daily grid box averages of 1-day point precipitation amounts, which is how one would interpret model output. That is, there is an order-of-operations difference that could potentially influence model-observations comparisons. If we had dense observational coverage (e.g., say at least 50 rain gauges in each model grid square), we could presumably first calculate the grid-box average precipitation amount, and then calculate Rx1day from those averages, bringing us conceptually closer to the thing that the models produce.

163-165: The paper seems to be getting ahead of itself a bit here. The results show this to be true for the precipitation variables you consider, but at this stage, it isn’t known whether we should expect the same for precipitation as for surface air temperature.

174: It seems clear from Fig. S2 that you have two groups of models with a substantial gap in the range of sensitivities that isn’t sampled. It would be worth including some discussion of how that heterogeneity could potentially have affected results.

200: Is the cost function minimized separately at each location? In minimizing the cost function, how do you account for the impacts of spatial and temporal dependence?
269-271: Since the fingerprints are model based, they could presumably be shown for the global domain, with outlines of the regions with observational coverage. But perhaps this suggestion is naïve (see my previous comment). If the cost function is minimized “globally” across the observation grid, then a question that emerges would be whether the geometry of the observation grid affects the fingerprints that are obtained, and what implications that might have for detection (or the timing of detection).

308-309: I think this is a consequence of having a collection of models that do not sample the model sensitivity spectrum very well. (See also my comment concerning line 174).

360: Confidence in the consistency of results perhaps? I’m not sure what it means for a method to be consistent.

387: I’m not sure I understand what it means for change to be in phase, but of opposite sign.

393: Replace “… coefficients flip sign.” with “… coefficients that flip sign.”

495-496: There are certainly papers that use precipitation datasets with greater coverage, but it takes a lot of work and personal contacts to collect that data and organize it for use in a D&A study. But even with that kind of work, coverage over land is still not very good. I wonder if we will get to the stage where we might have enough trust in reanalyzed precipitation to use it in a D&A study?

Hide

RR by Anonymous Referee #2 (09 Dec 2022)

ED: Publish subject to minor revisions (review by editor) (09 Dec 2022) by Gabriele Messori

AR by Iris de Vries on behalf of the Authors (19 Dec 2022) Author's response Author's tracked changes Manuscript

ED: Publish as is (22 Dec 2022) by Gabriele Messori

AR by Iris de Vries on behalf of the Authors (29 Dec 2022) Manuscript

Journal article(s) based on this preprint

26 Jan 2023

| Highlight paper

Robust global detection of forced changes in mean and extreme precipitation despite observational disagreement on the magnitude of change

Iris Elisabeth de Vries, Sebastian Sippel, Angeline Greene Pendergrass, and Reto Knutti

Earth Syst. Dynam., 14, 81–100, https://doi.org/10.5194/esd-14-81-2023,https://doi.org/10.5194/esd-14-81-2023, 2023

Short summary Chief editor

Iris Elisabeth de Vries, Sebastian Sippel, Angeline Greene Pendergrass, and Reto Knutti

Supplement

https://doi.org/10.5194/egusphere-2022-568-supplement

Iris Elisabeth de Vries, Sebastian Sippel, Angeline Greene Pendergrass, and Reto Knutti

Viewed

Total article views: 501 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
324	160	17	501	36	3	6

HTML: 324
PDF: 160
XML: 17
Total: 501
Supplement: 36
BibTeX: 3
EndNote: 6

Views and downloads (calculated since 13 Jul 2022)

Month	HTML	PDF	XML	Total
Jul 2022	116	38	7	161
Aug 2022	52	22	0	74
Sep 2022	68	41	6	115
Oct 2022	30	24	3	57
Nov 2022	21	15	0	36
Dec 2022	25	16	1	42
Jan 2023	12	4	0	16
Feb 2023	0
Mar 2023	0
Apr 2023	0
May 2023	0
Jun 2023	0
Jul 2023	0
Aug 2023	0
Sep 2023	0
Oct 2023	0
Nov 2023	0
Dec 2023	0
Jan 2024	0
Feb 2024	0
Mar 2024	0
Apr 2024	0
May 2024	0
Jun 2024	0
Jul 2024	0
Aug 2024	0

Cumulative views and downloads (calculated since 13 Jul 2022)

Month	HTML	PDF	XML	Total
Jul 2022	116	38	7	161
Aug 2022	52	22	0	74
Sep 2022	68	41	6	115
Oct 2022	30	24	3	57
Nov 2022	21	15	0	36
Dec 2022	25	16	1	42
Jan 2023	12	4	0	16
Feb 2023	0
Mar 2023	0
Apr 2023	0
May 2023	0
Jun 2023	0
Jul 2023	0
Aug 2023	0
Sep 2023	0
Oct 2023	0
Nov 2023	0
Dec 2023	0
Jan 2024	0
Feb 2024	0
Mar 2024	0
Apr 2024	0
May 2024	0
Jun 2024	0
Jul 2024	0
Aug 2024	0

Viewed (geographical distribution)

Total article views: 452 (including HTML, PDF, and XML) Thereof 452 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 01 Sep 2024

Download

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Preprint (3433 KB)
Metadata XML

Short summary

Precipitation changes are an important consequence of climate change. We use a method based on models and observations to detect changes in mean and extreme precipitation caused by external influence. We detect forced change in three observational datasets for global mean and extreme precipitation, but different observational datasets show different magnitudes of forced change. It is thus uncertain how large the change is, which is important for prediction of future changes and adaptation.


Total:	0
HTML:	0
PDF:	0
XML:	0