Skip to main content
ARS Home » Research » Publications at this Location » Publication #155188
Title: A MULTI-SCALE REMOTE SENSING MODEL FOR DISAGGREGATING REGIONAL FLUXES TO MICROMETEOROLOGICAL SCALES

Author
item ANDERSON, MARTHA - UNIVERSITY OF WI
item NORMAN, JOHN - UNIVERSITY OF WI
item MECIKALSKI, JOHN - UNIVERSITY OF WI
item TORN, RYAN - UNIVERSITY OF WA
item Kustas, William - Bill
item BASARA, JEFFREY - OKLAHOMA CLIMATE SURVEY

Submitted to: Journal of Hydrometeorology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/15/2003
Publication Date: 4/5/2004
Citation: Anderson, M.C., Norman, J.N., Mecikalski, J.R., Torn, R.D., Kustas, W.P., Basara, J.B. 2004. A multi-scale remote sensing model for disaggregating regional fluxes to micrometeorological scales. Journal of Hydrometeorology. 5:343-363.

Interpretive Summary: A procedure for disaggregating regional evapotranspiration (ET) estimates from an operational remote sensing satellite platform and modeling system to the field scale was evaluated in Oklahoma during 2000-2001 growing season. A technique that sharpens surface temperature to higher pixel resolutions was also incorporated providing ET estimates at the same resolution as remotely sensed vegetation cover. The accuracy and utility of this combined multi-scale modeling system is evaluated quantitatively in comparison with measurements made with ET towers in the Oklahoma Mesonet, and qualitatively, in terms of enhanced information content that emerges at high resolution where flux patterns can be identified with recognizable surface phenomena. Disaggregated ET fields at high resolution were reaggregated over an area representative of the tower ET measurements, and agreement was to within 10%. In contrast, regional ET predictions from the operational system showed a higher relative error of nearly 20% due to the gross mismatch in scale between model and measurement, highlighting the efficacy of disaggregation as a means for validating regional-scale ET predictions over heterogeneous landscapes. Sharpening the surface temperature inputs significantly improved output in terms of visual information content and model convergence rate.

Technical Abstract: Disaggregation of regional-scale (103 m) flux estimates to micrometeorological scales (101-102 m) facilitates direct comparison between land-surface models and ground-based observations. Inversely, it also provides a means for upscaling flux tower information into a regional context. The utility of the Atmosphere Land-Exchange Inverse (ALEXI) model and associated disaggregation technique (DisALEXI) in effecting regional to local downscaling is demonstrated in an application to thermal imagery collected with the GOES (5-km resolution) and Landsat (60-m) satellites over the state of Oklahoma on four days during 2000-2001. A related algorithm (DisTrad) sharpens thermal imagery to resolutions associated with visible/near infrared bands (30 m on Landsat), extending the range in scales achievable through disaggregation. The accuracy and utility of this combined multi-scale modeling system is evaluated quantitatively in comparison with measurements made with flux towers in the Oklahoma Mesonet, and qualitatively, in terms of enhanced information content that emerges at high resolution where flux patterns can be identified with recognizable surface phenomena. Disaggregated flux fields at 30-m resolution were reaggregated over an area approximating the tower flux footprint, and agreed with observed fluxes to within 10%. In contrast, 5-km flux predictions from ALEXI showed a higher relative error of 17% due to the gross mismatch in scale between model and measurement, highlighting the efficacy of disaggregation as a means for validating regional-scale flux predictions over heterogeneous landscapes. Sharpening the thermal inputs to DisALEXI with DisTrad did not improve agreement with observations in comparison with a simple bilinear interpolation technique because the sharpening interval associated with Landsat (60 to 30 m) was much smaller than the dominant scale of heterogeneity (200-500 m) in the scenes studied. Greater benefit is expected in application to MODIS data, where the potential sharpening interval (1 km to 250 m) brackets the field scale. Thermal sharpening did, however, significantly improve output in terms of visual information content and model convergence rate.