Surface Water Extent Estimation using Stream Gage Data in Coastal Watersheds of Connecticut
Abstract
Coastal areas are increasingly vulnerable to flooding. Rising sea levels and changes in precipitation patterns threatens a higher frequency and a larger extent of water overflow affecting housing, infrastructure and water quality. Developing the statistics of flooding events, and more broadly of water extent variability in coastal areas would be an important step to understand the interaction of river discharge, surges and tides processes, and to anticipate its impacts. A major constraint to develop the statistics is the incomplete time-series of high-resolution water extent maps in the satellite imagery archives. Not only the overpass is low frequency, but also missing images of water extent due to cloud cover during acquisition times, contribute to the gaps. We seek to fill those gaps in the time series through enhanced relationships of maps with a combination of gages within corresponding river basins. Our analysis is performed for inland river basins and for coastal watersheds. Monthly maximum surface water extent (monthly water) is retrieved from the European Commission's Joint Research Centre (JRC) Monthly Water History from 1984 to 2018. Monthly water is extracted from Landsat satellite imageries using several statistical algorithms to detect maximum surface water in each month from 1984 to 2018 and is available on a monthly basis. Gage discharges are retrieved from U.S. Geological Survey National (USGS) Water Information System (NWIS). USGS NWIS provides instantaneous discharge, daily mean discharge, and statistics in daily, monthly, and annual. To pair the gage discharge with the monthly water, the monthly gage discharge data are categorized into two types of data: monthly average gage discharge and monthly maximum instantaneous gage discharge. As expected, results indicate coastal watersheds have a more complex relationship gage-water compared to inland watersheds. In the former case, the distribution of surface water extent shows a mixed-Gaussian distribution, which can be separated into well-defined clusters. In the latter case, a linear relationship gage-water is found. Our study indicates the necessity of analyzing coastal watersheds in combination with statistics of surges and tides. Various statistical methods to evaluate a hierarchy of processes are needed to generalize those relationships.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2020
- Bibcode:
- 2020AGUFMOS007..08Y
- Keywords:
-
- 4217 Coastal processes;
- OCEANOGRAPHY: GENERAL;
- 4262 Ocean observing systems;
- OCEANOGRAPHY: GENERAL;
- 4513 Decadal ocean variability;
- OCEANOGRAPHY: PHYSICAL;
- 4556 Sea level: variations and mean;
- OCEANOGRAPHY: PHYSICAL