Mapping Drought Sensitivity of Ecosystem Functioning in Mountainous Watersheds: Spatial Heterogeneity and Geological-Geomorphological Control
Abstract
Mountainous watersheds in the Upper Colorado River Basin play a critical role in supplying water and nutrients to western North America. Ecosystem functioning in those regions - including plant dynamics and biogeochemical cycling - is known to be limited by water availability. Under the climate change, early snowmelt and increasing temperature are expected to intensify the drought conditions in early growing seasons. Although the impact of early-season drought has been documented in plot-scale experiments, ascertaining its significance in mountainous watersheds is challenging given the highly heterogeneous nature of the systems with complex terrain and diverse plant functional types (PFTs). The objectives of this study are (1) to map the regions where the plant dynamics are relatively more sensitive to drought conditions based on historical satellite and climate data, and (2) to identify the environmental controls (e.g., geomorphology, elevation, geology, snow and PFT) on drought sensitivity. We characterize the spatial heterogeneity of drought sensitivity in four watersheds (a 15 x 15 km domain) near the Rocky Mountain Biological Laboratory in Colorado, USA. Following previous plot-scale studies, we first define the drought sensitivity based on annual peak NDVI (Landsat 5) and climatic datasets. Non-parametric tree-based machine learning methods are used to identify the significant environmental controls, using high-resolution LiDAR digital elevation map and peak snow-water-equivalent distribution from NASA airborne snow observatory. Results show that the drought sensitivity is negatively correlated with elevation, suggesting increased water limitations in lower elevation (less snow, higher temperature). The drought sensitivity is more spatially variable in shallow-rooted plant types, affected by local hydrological conditions. We also found geomorphological and geological controls, such as high sensitivity in the steep well-drained glacial moraine regions. Our results highlight the importance of geology and subsurface flow conditions, in addition to snow accumulation. In parallel, the remotely-sensed drought sensitivity can be used as a scalable metric to identify the vulnerable regions to the future climate change, as well as to inform future sampling and characterization.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2016
- Bibcode:
- 2016AGUFM.H11D..06W
- Keywords:
-
- 0740 Snowmelt;
- CRYOSPHEREDE: 1621 Cryospheric change;
- GLOBAL CHANGEDE: 1813 Eco-hydrology;
- HYDROLOGYDE: 1817 Extreme events;
- HYDROLOGY