A parsimonious process-based regionalization of hydrologic response across Alaska
Abstract
In remote and mountainous regions, including much of Alaska, the paucity of streamflow and hydrometeorological measurements can lead to substantial uncertainties in the quantification of the water balance. Spatially distributed datasets from models and remote sensing are increasingly available and provide information that can be used to improve integrated hydrologic modeling within these regions. The U.S. Geological Survey is extending its National Hydrologic Model infrastructure to include Alaska and contiguous Canadian watersheds, hereafter the Alaska Domain (AD). The objective of this research was to evaluate the variability of hydrologic response across the AD using streamflow observations along with various spatially distributed datasets. We compiled streamflow observations for 84 watersheds that included a minimum record of 10 complete water years from 2000 - 2019 and quantified hydrologic response by calculating 7 annual hydrologic signatures that can be used to characterize the primary properties of the daily streamflow hydrograph. Watershed topography, climate, hydrology, snow and ice, aridity, and geology characteristics were also compiled from global modeling and remote sensing datasets. Furthermore, given the importance of snow and glaciers to hydrology across the AD, we developed sub-daily simulations at moderate (1 km) resolution for 2000 - 2019 using the SnowModel snow evolution modeling system. Random forest models were developed for each of the hydrologic signatures to evaluate the relative importance of watershed characteristics in explaining variability in hydrologic response. Glacial fraction was shown to be the most important variable explaining the amplitude, lag-1 autocorrelation, and kurtosis. The seasonality of snowmelt and rainfall input was most important for the coefficient of variation, skew, and phase. Lastly, aridity showed a dominant importance for the mean daily streamflow. Permafrost probability was shown to be an important secondary control on all hydrologic signatures. These results indicate the regional importance of hydrologic signatures based on interactions between three primary process-based drivers: aridity, input seasonality, and glacial fraction, and can be used to help guide hydrologic modeling development across the AD.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2020
- Bibcode:
- 2020AGUFMH014...04S
- Keywords:
-
- 1825 Geomorphology: fluvial;
- HYDROLOGY;
- 1855 Remote sensing;
- HYDROLOGY;
- 1856 River channels;
- HYDROLOGY;
- 1860 Streamflow;
- HYDROLOGY