An initial validation of Landsat 5 and 7 derived surface water temperature for U.S. lakes, reservoirs, and estuaries. Issue 22 (17th November 2018)

• Record Type: Journal Article
• Authors: Schaeffer, Blake A.; Iiames, John; Dwyer, John; Urquhart, Erin; Salls, Wilson; Rover, Jennifer; Seegers, Bridget
• Is Part Of: International journal of remote sensing. Volume 39:Issue 22(2018)
• Page Start: 7789
• Page End: 7805
• Publication Date: 2018-11-17
• Subjects: Remote sensing -- Periodicals; Télédétection -- Périodiques; 621.3678
• DOI: 10.1080/01431161.2018.1471545
• Languages: English
• ISSNs: 0143-1161

Abstract

ABSTRACT: The United States Harmful Algal Bloom and Hypoxia Research Control Act of 2014 identified the need for forecasting and monitoring harmful algal blooms (HAB) in lakes, reservoirs, and estuaries across the nation. Temperature is a driver in HAB forecasting models that affects both HAB growth rates and toxin production. Therefore, temperature data derived from the U.S. Geological Survey Landsat 5 Thematic Mapper and Landsat 7 Enhanced Thematic Mapper Plus thermal band products were validated across 35 lakes and reservoirs, and 24 estuaries. In situ data from the Water Quality Portal (WQP) were used for validation. The WQP serves data collected by state, federal, and tribal groups. Discrete in situ temperature data included measurements at 11, 910 U.S. lakes and reservoirs from 1980 through 2015. Landsat temperature measurements could include 170, 240 lakes and reservoirs once an operational product is achieved. The Landsat-derived temperature mean absolute error was 1.34°C in lake pixels >180 m from land, 4.89°C at the land-water boundary, and 1.11°C in estuaries based on comparison against discrete surface in situ measurements. This is the first study to quantify Landsat resolvable U.S. lakes and reservoirs, and large-scale validation of an operational satellite provisional temperature climate data record algorithm. Due to the high performance of open water pixels, Landsat satellite data may supplement traditional in situ sampling by providing data for most U.S. lakes, reservoirs, and estuaries over consistent seasonal intervals (even with cloud cover) for an extended period of record of more than 35 years.