Effect of Landsat Data Pre-processing on Reservoir Surface Area Extraction
Abstract
Availability of no-cost Landsat data have resulted in increased number of land cover mapping and monitoring applications. Extracting surface area of water bodies and estimating changes over time is a common application. Landsat data are pre-processed to multiple levels and are available through few different sources. For example, USGS provides Landsat data as collections and users can access them as digital numbers, top of atmosphere (collection 1 only) or surface reflectance products. Values recorded by pixels are altered during each pre-processing step and impact the outcome of a mapping and monitoring project. In this study, we analyzed 3-levels of pre-processed Landsat data for extracting surface area values of reservoirs located along a latitudinal gradient in MT, WY, CO, and NM. Two widely used water indices were used for quantifying the effect of pre-processing. These results highlight systematic differences in the surface area values depending on the type of pre-processed Landsat data. Surface area estimates can be systematically biased and introduce uncertainties in models that use remotely sensed data for water resource management.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2022
- Bibcode:
- 2022AGUFM.H25T1353S