Inferring anthropogenic trends from satellite data for water-sustainability of US cities near artificial reservoirs

- We analyzed trend and relationship between multiple satellite and ground data.
- We used precipitation, NDVI, land surface temperature, land cover, and streamflow.
- Streamflow, precipitation, NDVI has been decreasing with increasing temperature.
- Temperature projections can be used in understanding future water availability.
- Larger watersheds are potentially more sustainable and resilient to climate change.

Anthropogenic activities affect the water cycle and water supply at global and regional spatial scales, and approaches to water management must consider anthropogenic inputs. One of the major inputs in local-to-regional availability of water and the water cycle is land use land cover change as a result of urbanization, artificial reservoirs, and irrigation activity. To understand evolving trends in local hydrologic cycle for water sustainability of growing cities, this study employed a multi-factorial approach involving population trends, water use (and demand), streamflow, and various satellite-derived water-relevant variables. These variables are daily precipitation (from Tropical Rainfall Measuring Mission-TRMM, 3B42.V7), Normalized Difference Vegetation Index (NDVI) (from Moderate Resolution Imaging Spectroradiometer-MODIS-MOD13A1), land surface temperature (LST) (from MODIS-MOD11A2), and land cover (MODIS-MCD12Q1). Long term trends in such data were used to understand temporal and spatial trends in impounded watersheds hosting a large and growing city. The cities studied for water sustainability were Atlanta, Georgia and Buford dam; Columbia, South Carolina and Saluda dam; Columbus, Ohio and Alum Creek dam; Montgomery, Alabama and Jordan dam; Tulsa, Oklahoma and Keystone dam; and Tuscaloosa, Alabama and Tuscaloosa dam. Our study reveals that daily mean stream flow has been decreasing in all but one (Tulsa) of the areas selected. Satellite data trends between 2000 and 2012 showed a steady decrease in precipitation and NDVI, while LST has gradually increased. We attribute the NDVI (i.e., gradual decrease in vegetation cover) to LST rather than precipitation trends. The results of this research suggest that future temperature projections from climate models can be used in understanding vegetation activity and water availability over the study areas. Cities with larger upstream watershed area are potentially more sustainable and resilient (than those with small watersheds) as a result of spatial variability of water resources' response to climate change. Inter-basin water resources transfer is a possible solution to vulnerable cities in the future. The study results also emphasize the need to establish a sustainable and resilient water resources management system that includes narrowing the information and perception gap between the engineering community and the general public.