MONITORING SPATIOTEMPORAL TOTAL ORGANIC CARBON CONCENTRATIONS IN LAKE MEAD WITH INTEGRATED DATA FUSION AND MINING (IDFM) TECHNIQUE

摘要

Forest fires, soil erosion, and land use changes in watersheds nearby Lake Mead and inflows from Las Vegas Wash into the lake are considered as sources of the lake's water quality impairment. These conditions result in higher concentration of Total Organic Carbon (TOC). TOC in contact with Chlorine which is often used for disinfection purposes of drinking water supply causes the formation of trihalomethanes (THMs). THM is one of the toxic carcinogens controlled by the EPA's disinfection by-product rule. As a result of the threat posed to drinking water used by the 25 million people downstream, recreational area, and wildlife habitat of Lake Mead, it is necessary to develop a method for near real-time monitoring of TOC in this area. Monitoring through a limited number of ground-based monitoring stations on a weekly/monthly basis is insufficient to capture both spatial and temporal variations of water quality changes. In this study, remote sensing technology with the aid of data fusion and mining techniques provides us with information about the spatiotemporal distribution of TOC for the entire lake on a daily basis. A data fusion method was applied to bridge the gap of poor 250/500m spatial resolution for the land bands of Moderate Resolution Imaging Spectroradiometer (MODIS) imageries with the 30 m enhanced spatial resolution of Landsat's imageries which suffers from long overpass of 16 days. Consequently, Integrated Data Fusion and Mining (IDFM) techniques produce synthetic fused images of MODIS and Landsat satellites with both high spatial and temporal resolution to create near real time TOC distribution maps and lead to sustainable water quality management with the aid of IDFM in Lake Mead watershed.