Casi-atm Observed and Simulated Ecohydrological Relevant Water and Energy Fluxes

Monitoring of water and energy fluxes is a requirement for assessment of climate and anthropogenic effects on natural and agricultural ecosystems [i]. These fluxes consist of the physical transfers between soil, vegetation and atmosphere. In order to describe these transfers, on local, regional or global scale, and the changes taking place in the compartments it is necessary to have measurements and models to estimate these fluxes. Distributed hydrological models are appropriate tools and are constantly improving the understanding of this cycling of water through soils, vegetation and atmosphere. Also for the relatively new emerging field of ecohydrology estimation of spatially distributed water and energy levels is essential for predicting occurrence of vegetation species under site specific conditions. Since evapotranspiration is a major component of all combined fluxes it deserves to receive the necessary attention. For the estimation of remotely sensed evapotranspiration, the most popular approach is to use surface temperature and vegetation indices at small scale and low resolution satellite imagery to estimate regional fluxes. New developments in thermal airborne (ATM) combining with CASI imagery allow deriving hydrological relevant observations at a resolution of 1-10 m. In this research we propose to estimate on basis of thermal ATM imagery evapotranspiration and other elements of the water and energy balance on a scale, which allows discriminating local wetness and vegetation heterogeneity in relation to differences in soil and vegetation condition. The evapotranspiration in the study area is simulated with the Surface Energy Balance Algorithm for Land (SEBAL) [ii]. The daily total evapotranspiration is calculated for the complete study area, but special attention is given to a grassland area for which detailed soil moisture measured conditions have been obtained. The estimated evapotranspiration values can be related to other hydrological and vegetation conditions, resulting in an increased understanding of the ecohydrological functioning of the study area. INTRODUCTION The estimation of water balances at different spatial and temporal scales is a fundamental task of the hydrological science. In recent years physically based distributed models are more and more used for this task. These models simulate water and energy fluxes between soil, vegetation and atmosphere, their applications range from modeling climate change on global scale to ecohydrological modeling on local scale. Characteristic is that they require spatially distributed data and that the evapotranspiration is general the biggest flux to be simulated. The Surface Energy Balance Algorithm for Land (SEBAL) [ii] model is an image-processing and GIS model comprised of 25 computational steps that calculate the actual (ETact) and potential evapotranspiration rates (ETpot) through the balance of sun energy on the earth surface. SEBAL has been so far mainly been applied on regional scale on the basis of NOAA-AVHRR imagery. The key input data for SEBAL consists of spectral radiance in the visible, near-infrared and thermal infrared part of the spectrum. SEBAL computes a complete radiation and energy balance along with the resistances for momentum, heat and water vapour transport. The resistances are a function of state conditions such as soil water potential, wind speed and air temperature and change from day-to-day. © EARSeL and Warsaw University, Warsaw 2005. Proceedings of 4th EARSeL Workshop on Imaging Spectroscopy. New quality in environmental studies. Zagajewski B., Sobczak M., Wrzesień M., (eds) In this paper we test the applicability of the SEBAL methodology for estimation of evapotranspiration fluxes with spatially high resolution hyperspectral imagery (CASI) and thermal data from the ATM sensor. SURFACE ENERGY BALANCE ALGORITHM FOR LAND The primary basis for the SEBAL model is the surface energy balance (Fig. 1). The instantaneous ETact flux is calculated for each pixel of the image as a ‘residual’ of the surface energy budget equation Figure 1: Surface energy balance [ii]. ET=Rn-G-H (1) where ET is the latent heat flux [W/m], Rn is the net radiation flux at the surface [W/m], G is the soil heat flux [W/m], and H is the sensible heat flux to the air [W/m]. Rn represents the actual radiant energy available at the surface. It is computed by subtracting all outgoing radiant fluxes from all incoming radiant fluxes (Fig. 2). This is specified in the surface radiation balance equation, ↓ ↑ ↓ ↓ ↓ − − − + − = L L L s s n R R R R R R ) 1 ( 0 ε α (2) where RS↓ is the incoming short-wave radiation [W/m], α is the surface albedo [-], RL↓ is the incoming long wave radiation [W/m], RL↑ is the outgoing long wave radiation [W/m], and εo is the surface thermal emissivity [-]. In Eq. (2) the amount of net short-wave radiation (RS↓ αRS↓) that remains available at the surface is a function of the surface albedo (α). The broadband surface albedo α is derived from the narrow band spectral reflectances α(λ) measured by each satellite band. The incoming short-wave radiation (RS↓) is computed using the solar constant, the solar incidence angle, the relative earth-sun distance and a computed broadband atmospheric transmissivity. This latter transmissivity can be estimated from sunshine duration or inferred from pyranometer measurements (if available). The incoming long wave radiation (RL↓) is computed using a modified Stefan-Boltzmann equation with an apparent emissivity that is coupled to the shortwave atmospheric transmissivity and a measured air temperature. Outgoing long wave radiation (RL↑) is computed using the Stefan-Boltzmann equation with a calculated surface emissivity and surface temperature. Surface temperatures are derived from the satellite measurements of thermal radiances. In Eq. (1), the soil heat flux (G) and sensible heat flux (H) are subtracted from the net radiation flux at the surface (Rn) to compute the "residual" energy available for evapotranspiration (λE). Soil heat flux is empirically calculated as a G/Rn fraction using vegetation indices, surface temperature, and surface albedo. Sensible heat flux is computed using wind speed observations, estimated surface roughness, and surface to air temperature differences that are obtained through a sophisticated self-calibration between dry (λE≈0) and wet (H≈0) pixels. SEBAL uses an iterative process to correct for atmospheric instability caused by buoyancy effects of surface heating. The λE time integration in SEBAL is split into two steps. The first step is to convert the instantaneous latent heat flux into daily λE24 values by holding the evaporative fraction constant. The evaporative fraction, EF [-], is: EF = λE/ (Rn G) (3)