Comparison of Land Cover Information from LANDSAT MSS and Airborne TMS for Hydrological Applications: Preliminary Results
J. C. Gervin, Y. C. Lu, W. A. Hallada, R. F. Marcell
Abstract
J. C. Gervin, Y. C. Lu, W. A. Hallada, R. F. Marcell
Abstract
Land cover information for the Clinton River Basin (Michigan) derived from LANDSAT multispectral scanner (MSS) data was compared with that from airborne thematic mapper simulator (TMS) to investigate the probable capabilities of the thematic mapper (TM) launched aboard LANDSAT-4 in July 1982. The preliminary findings for one 7.5 minute topographic map, Mt. Clemens West, are reported. Significant improvements in land cover classification accuracy were obtained using TMS data as compared with MSS data. Overall mapping accuracy increased from 49 to 61 percent with an improvement from 71 to 84 percent in the residential category. A combination of four bands with one band in each major region of the spectrum (visible, near IR, middle IR and thermal IR) provided as good a discrimination of land cover as all seven TM bands. Based on the improved land cover classification accuracy of TM, TM data has the potential to provide more useful and effective input to US Army Corps of Engineers flood forecasting and flood damage prediction/assessment models.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Land cover information for the Clinton River Basin (Michigan) derived from LANDSAT multispectral scanner (MSS) data was compared with that from airborne thematic mapper simulator (TMS) to investigate the probable capabilities of the thematic mapper (TM) launched aboard LANDSAT-4 in July 1982. The preliminary findings for one 7.5 minute topographic map, Mt. Clemens West, are reported. Significant improvements in land cover classification accuracy were obtained using TMS data as compared with MSS data. Overall mapping accuracy increased from 49 to 61 percent with an improvement from 71 to 84 percent in the residential category. A combination of four bands with one band in each major region of the spectrum (visible, near IR, middle IR and thermal IR) provided as good a discrimination of land cover as all seven TM bands. Based on the improved land cover classification accuracy of TM, TM data has the potential to provide more useful and effective input to US Army Corps of Engineers flood forecasting and flood damage prediction/assessment models.
Key concepts: Thematic Mapper, Multispectral Scanner, Land cover, Remote sensing, Flood myth, Multispectral image, Environmental science, Thermal infrared