The JORNada EXperiment (JORNEX) held at the Jornada Experimental Range in southern New Mexico aims at the description of the surface energy balance of a desert grassland ecosystem. A large array of both field and remote sensing data have been collected from 1995 to 1998. Remote sensing data collected consists of visible and thermal infrared imagery, video imagery and laser profiler data. These data are being used to infer land surface characteristics such as surface temperature, albedo and roughness. These land surface characteristics can be used as input for surface energy balance models coupled with atmospheric models. However atmospheric models work with very coarse grid cells of at least 50*50 km. Remote sensing data typically has a geometrical resolution much smaller than 50*50 km. In addition to the scale discrepancy, surface energy balance models are generally non-linear algorithms, which amplifies scale problems. High resolution input data will not necessarily give the same results as low resolution input data. The problem arises of how to scale up the calculated surface energy balance fluxes to the much coarser resolution of the atmospheric model. The first step is to derive the length scales of the land surface characteristics involved in the calculation of the surface energy balance, therefore determining the maximum resolution required to sample the area. High resolution (~4m) NIR and TIR images with variogram and wavelet techniques will be used to study scaling.