the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
The influence of permafrost and other environmental controls on stream thermal sensitivity across Yukon, Canada
Abstract. Thermal sensitivity, defined as the slope of a linear regression between stream and air temperature, is a useful indicator of the strength of coupling between atmospheric forcings and stream temperature, or conversely, of the presence of non-atmospheric thermal influences such as groundwater contributions to streamflow. Furthermore, thermal sensitivity is known to be responsive to environmental change. This study expands the current state of knowledge of stream thermal sensitivity in cold, northern regions across catchment scales, investigates the environmental controls of thermal sensitivity across a range of catchment dispositions, and assesses the thermal influence of environmental conditions unique to cold regions, namely permafrost. We conducted a linear regression analysis relating mean daily air and stream temperature in 57 catchments in Yukon, Canada, with catchment areas ranging from 5.4 to 86,500 km2, and with catchment mean permafrost probabilities ranging from 0.0 to 0.99. Thermal sensitivities obtained from the linear regressions ranged from 0.14 to 0.84 °C °C-1, with a median of 0.56 °C °C-1, and the regression intercepts ranged from -0.07 to 7.60 °C, with the mean regression Nash-Sutcliffe efficiency = 0.81. Thermal sensitivity was positively related to catchment area, land covers representing surface water storage, and streamflow ‘flashiness’ or a lack of groundwater contributions. The greatest single environmental characteristic explaining the variance in thermal sensitivity was catchment topography and physiography (9 % variance explained); however, 39 % of the variance in thermal sensitivity was explained jointly by catchment physiography, land cover, and permafrost presence indicators, suggesting thermal sensitivity is the result of multiple interacting controls. Permafrost appeared to have indirect and offsetting effects on thermal sensitivity through its influence on separate and counter-acting processes controlling thermal sensitivity.
- Preprint
(2195 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2024-1741', Anonymous Referee #1, 10 Sep 2024
General Comments
This manuscript was well written overall and addressed an interesting question of which factors have the greatest controls on stream thermal sensitivity in northern regions containing some amount of permafrost. The study uses linear regression and related statistical methods to assess the strength of different parameters’ influence on the thermal sensitivity. The scientific significance was good, with the scope very relevant to the special issue the manuscript has been submitted to. While the methods are not super novel, the authors have investigated a region (northern latitude, Yukon Canada) for which stream thermal sensitivity is much less studied since much research has been done on more temperature regions. The scientific quality also seems good, though I am not an expert in redundancy analysis and so comments from another reviewer on this aspect of the manuscript would be helpful. Lastly, the presentation quality is also good, with clear figures and tables and well written text throughout the manuscript.
Specific Comments
Line 426-437: Why does your model (and other models) over-predict during August – October? Any hypotheses as to why there is this seasonal variability? What is it about the permafrost catchments that makes their deviation over time greater?
Figures: There are 10 figures. Could some figures be consolidated/combined or placed in a supplement?
Figure 7 is a hard to interpret. Please provide a description either in the caption or in the text to help the reader interpret.
Figure 9 is also hard to interpret. Could you provide an explanation of how to interpret? Why do some of the regions in the diagram not have a value? Because they are negative?
Citation: https://doi.org/10.5194/egusphere-2024-1741-RC1 -
AC1: 'Reply on RC1', Andras Szeitz, 23 Oct 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-1741/egusphere-2024-1741-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Andras Szeitz, 23 Oct 2024
-
RC2: 'Comment on egusphere-2024-1741', Anonymous Referee #2, 30 Sep 2024
General Comments:
The manuscript The influence of permafrost and other environmental controls on stream thermal sensitivity across Yukon, Canada is a well written and important work on the patterns of correlation between modeled air and observed mean daily water temperatures in northern watersheds. The authors provide multiple analyses that demonstrate the complex relationship between air temperature, water temperature, and environmental conditions and make a compelling case that permafrost may mediate the relationship between air and water temperatures via controls on residence time and sub-surface flow paths.
Specific Comments:
I provide numerous suggestions for potential ways to strengthen the manuscript in the attached pdf for the authors to consider and highlight two comments below.
First, I suggest a discussion of sources of uncertainty and their implications on the outcome be added to the manuscript. This includes the use of modeled air temperature, the reporting uncertainty of water temperature observations on the order of 0.4 degrees C, etc.
Second, we know that this system is dynamic and in flux. As such, the thermal processes that drive correlations between air and water temperature are likely also changing over time. I understand the need to use static estimates for active layer thickness and permafrost coverage (limited by data), but that does assume static conditions (over 20+ years in some cases). This should be acknowledged in the discussion and the implications explained. The more interesting/concerning pattern that emerges from the data (which the authors are commended for publishing https://zenodo.org/records/11668943) is that thermal sensitivity varies substantially between years with a median interannual range (max TS – min TS) of 0.22, or 41% of the magnitude of the median TS value. To make it even more interesting, the interannual variability appears to be serially correlated (e.g., demonstrating a temporal trend) with 10 stations decreasing the TS over time and 2 increasing with a p value of 0.1 (see attached figure on the last page of “egusphere-2024-1741_review_and_figure.pdf”). The authors need not add this analysis to their work to make a sound contribution, but the potential (and apparent reality) on non-stationarity should be acknowledged and discussed.
Technical Corrections:
See attached document (egusphere-2024-1741_review_and_figure.pdf) for specific technical comments.
-
AC2: 'Reply on RC2', Andras Szeitz, 23 Oct 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-1741/egusphere-2024-1741-AC2-supplement.pdf
-
AC3: 'Reply on RC2', Andras Szeitz, 23 Oct 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-1741/egusphere-2024-1741-AC3-supplement.pdf
-
AC4: 'Reply on RC2', Andras Szeitz, 23 Oct 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-1741/egusphere-2024-1741-AC4-supplement.pdf
-
AC2: 'Reply on RC2', Andras Szeitz, 23 Oct 2024
Data sets
Research data set Andras J. Szeitz https://doi.org/10.5281/zenodo.11527471
Viewed
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
234 | 86 | 122 | 442 | 14 | 16 |
- HTML: 234
- PDF: 86
- XML: 122
- Total: 442
- BibTeX: 14
- EndNote: 16
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1