A data-model fusion approach to identify and map zones of similar hydrologic function across an alpine catchment
Abstract
Climate warming is changing alpine hydrologic processes, including precipitation patterns and the timing of snowmelt, which influence plant ecology, biogeochemistry, and water supply provisions to lower elevations. An important step towards predicting future changes in the alpine is to understand how the physical structure of the critical zone (CZ) determines patterns of water storage, routing, and export at the catchment scale. In this study, we used a data-model fusion approach to identify and predict areas with similar temporal patterns in soil water storage across a 0.26 km2 alpine catchment in the Colorado Rocky Mountains, U.S.A. We first used an unsupervised hierarchical cluster analysis to identify six unique temporal patterns in soil moisture observations, ranging from predominately dry to persistently very wet. We then explored relationships among the six hydrologic "functions" and multiple spatially extensive CZ-related indices, including snow depth, plant productivity, macro- (102- >103 m) and microtopography (< 100-102 m), and flow paths. Finally, we used a supervised machine learning random forest algorithm to map each of the hydrologic groups into zones across the catchment and evaluated their aggregate relationships. Our analysis indicated that ~50% of the catchment is hydrologically connected to the channel, lending insight into the portions of the catchment that dominate stream water and solute fluxes. The drier half of the catchment is likely important for localized nutrient cycling and plant species diversity but has less influence on ecosystem- or catchment-scale processes. Although the wettest hydrologic groups were only ~5% of the total catchment area and had the lowest predictive accuracy, patches characterized by these groups may disproportionately contribute to biogeochemical processes and plant productivity, necessitating further study. Our approach demonstrated the utility of combining multiple scales of observations with statistical modeling to expand our understanding of patch-to-catchment-scale CZ controls on alpine hydrology, which can ultimately inform process-based models aimed at quantifying future changes in alpine ecosystems.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2020
- Bibcode:
- 2020AGUFMH072...05H
- Keywords:
-
- 0414 Biogeochemical cycles;
- processes;
- and modeling;
- BIOGEOSCIENCES;
- 1807 Climate impacts;
- HYDROLOGY;
- 1813 Eco-hydrology;
- HYDROLOGY;
- 1834 Human impacts;
- HYDROLOGY