Remote sensing of surface energy fluxes at 101-m pixel resolutionsJ. M. Norman,1M. C. Anderson,1W. P. Kustas,2A. N. French,2,3J. Mecikalski,4R. Torn,1,5G. R. Diak,4T. J. Schmugge,2and B. C. W. Tanner1,6Received 16 October 2002; revised 18 March 2003; accepted 7 May 2003; published 28 August 2003.[1] Many applications exist within the fields of agriculture, forestry, land management,and hydrologic assessment for routine estimation of surface energy fluxes, particularlyevapotranspiration (ET), at s patial resolutions of the order of 101m. A new two-stepapproach (called the disaggregated atmosphere land exchange inverse model, orDisALEXI) has been developed to combine low- and high-resolution remote sensing datato estimate ET on the 101–102m scale without requiring any local observations. The firststep uses surface brightness-temperature-change measurements made over a 4-hourmorning interval from the GOES satellite to estimate average surface fluxes on the scale ofabout 5 km with an algorithm known as ALEXI. The second step disaggregates the GOES5-km surface flux estimates by using high-spatial-resolution images of vegetationindex and surface temperature, such as from ASTER, Landsat, MODIS, or aircraft, toproduce high-spatial-resolution maps of surface fluxes. Using data from the SouthernGreat Plains field experiment of 1997, the root-mean-square difference between remoteestimates of surface fluxes and ground-based measurements is about 40 W m2,comparable to uncertainties associated with micrometeorological surface fluxmeasurement techniques. The DisALEXI approach was useful for estimating field-scale,surface energy fluxes in a heterogeneous area of central Oklahoma without using anylocal observations, thus providing a means for scaling kilometer-scale flux estimates downto a surface flux-tower footprint. Although the DisALEXI approach is promising forgeneral applicability, further tests with varying surface conditions are necessary toestablish greater confidence.INDEX TERMS: 3360 Meteorology and Atmospheric Dynamics:Remote sensing; 3322 Meteorology and Atmospheric Dynamics: Land/atmosphere interactions; 1818Hydrology: Evapotranspiration; 3337 Meteorology and Atmospheric Dynamics: Numerical modeling and dataassimilation; KEYWORDS: thermal infrared, remote sensing, disaggregation, surface flux modelsCitation: Norman, J. M., M. C. Anderson, W. P. Kustas, A. N. French, J. Mecikalski, R. Torn, G. R. Diak, T. J. Schmugge,and B. C. W. Tanner, Remote sensing of surface energy fluxes at 101-m pixel resolutions, Water Resour. Res., 39(8), 1221,doi:10.1029/2002WR001775, 2003.1. Introduction[2] The partitioning of net radiation at a land surface intolatent and sensible heat fluxes reflects the physical charac-teristics of the surface: its roughness, moisture content,vegetation cover, and so forth. Surface flux partitioningalso influences the coupling between the surface and thelower atmosphere. Whether one is interested in studying theinfluence of surface fluxes on the lower atmosphere or incharacterizing the nature of the surface itself, much can belearned by quantifying the temporal and spatial character ofthe surface energy balance using remote ly sense d data.There is particular demand for operational methodologiesfor mapping surface fluxes, algorithms that can be executedremotely and routinely and do not require any local, ground-based data as input.[3] Numerous applications exist for operational maps ofenergy balance/evapotranspiration (ET) made at a broadrange in spatial scales. The routine estimation of surfacefluxes at regional scales (i.e., 10–100 km), for example,would benefit numerical weather forecasting in terms ofdefining accurate model boundary conditions. At highspatia l resolution (i.e., 101m), remote estimates of ETprovide a means to calibrate models simulating site-specificenergy and water balances of agricultural crops, which usetranspiration-rate estimates to detect productivity-limitingfield conditions and project end-of-season yields [Moran etal., 1995; Moulin et al., 1998]. In forested areas, thetranspiration rate is also the primary indicator of foresthealth and vulnerability to fire [Vidal and Devaux-Ros,1995]. Changes in the energy balance are indicative ofchanges in cropland and natural ecosystem functioning;hence remote sensing models have potential in mappingand monitoring plant ecosystem health [Moran, 2003]. At1Department of Soil Science, University of Wisconsin, Madison,Wisconsin, USA.2USDA-ARS, Hydrology and Remote Sensing Lab, Beltsville, Mary-land, USA.3Now at Hydrological Sciences Branch, NASA Space Flight Center,Greenbelt, Maryland, USA.4Space Science and Engineering Center, University of Wisconsin,Madison, Wisconsin, USA.5Now at Department of Atmospheric Science, University of Washington,Seattle, Washington, USA.6Now at Biomedical Engineering, University of Washington, Seattle,Washington, USA.Copyright 2003 by the American Geophysical Union.0043-1397/03/2002WR001775$09.00SWC 9 - 1WATER RESOURCES RESEARCH, VOL. 39, NO. 8, 1221, doi:10.1029/2002WR001775, 2003these scales, maps of daily ET can also be used to constrainrecharge in detailed transient hydrologic models.[4] Unfortunately, methodologies using high-resolutionremotely sensed data with 101–102m pixel resolutiongenerally rely on the availability of contemporaneous insitu measurements, primarily near-surface meteorologicalconditions such as air temperature, wind speed, andhumidity, and are therefore difficult to implement opera-tionally [Gardner et al., 1992; Choudhury et al., 1994;Moran et al. , 1994; Moran et al., 1996]. Routine applica-tion of high-resolution satellite data is also hindered by thelong return period between successive satellite overpasses.The frequency of repeated coverage from the Land Re-mote-Sensing Satellite (Landsat) or the Advanced Space-borne Thermal Emission Reflectance Radiometer(ASTER), for example, is typically of the order of severalweeks. Considering that cloud cover will obscure somefraction of these images, monthly coverage is a reasonableexpectation for the availability of high-resolution satelliteimages. Using one snapshot of the surface per month forestimating spatially distributed surfac e heat fluxes andevapotranspiration (ET) severely limits the utility foroperational monitoring of vegetation conditions [Moranet al., 1996].[5] For coarser resolutions of 103–104m, satellite cover-age is much more frequent, and hence surface fluxes havebeen