Synopsis
This user’s guide describes a level 4 MODIS land data product, MOD16, the global 8-day (MOD16A2) and annual (MOD16A3) terrestrial ecosystem Evapotranspiration (ET) dataset at 0.5 km spatial resolution over the 109.03 Million km2 global vegetated land areas designed for the MODIS sensor aboard the Aqua and Terra platforms. The MOD16 algorithm is based on the logic of the Penman-Monteith equation which uses daily meteorological reanalysis data and 8-day remotely sensed vegetation property dynamics from MODIS as inputs. This document is intended to provide both a broad overview and sufficient detail to allow for the successful use of the data in research and applications.
Please note the “MOD” prefix should be considered as referring to data sets derived from MODIS onboard either TERRA or Aqua satellite. That is, “MOD” in this document can also be treated as “MYD” derived from MODIS on Aqua.
1. The Algorithm, Background and Overview
Calculation of ET is typically based on the conservation of either energy or mass, or both. Computing ET is a combination of two complicated major issues: (1) estimating the stomatal conductance to derive transpiration from plant surfaces; and (2) estimating evaporation from the ground surface. The MOD16 ET algorithm runs at daily basis and temporally, daily ET is the sum of ET from daytime and night. Vertically, ET is the sum of water vapor fluxes from soil evaporation, wet canopy evaporation and plant transpiration at dry canopy surface. Remote sensing has long been recognized as the most feasible means to provide spatially distributed regional ET information on land surfaces. Remotely sensed data, especially those from polar-orbiting satellites, provide temporally and spatially continuous information over vegetated surfaces useful for regional measurement and monitoring of surface biophysical variables affecting ET, including albedo, biome type and leaf area index (LAI) (Los et al., 2000).
Energy partitioning at the surface of the earth is governed by the following three coupled equations:
where H, E and A′ are the fluxes of sensible heat, latent heat and available energy for H and E; Rnet is net radiation, G is soil heat flux; ∆S is the heat storage flux. is the latent heat of vaporization. is air density, and CP is the specific heat capacity of air; Ts, Ta are the aerodynamic surface and air temperatures; ra is the aerodynamic resistance; esat, e are the water vapour pressure at the evaporating surface and in the air; rs is the surface resistance to evapotranspiration, which is an effective resistance to evaporation from land surface and transpiration from the plant canopy. The psychrometric constant is given by
where Ma and Mw are the molecular masses of dry air and wet air respectively and Pa the atmospheric pressure.
Developing a robust algorithm to estimate global evapotranspiration is a significant challenge. Traditional energy balance models of ET require explicit characterization of numerous physical parameters, many of which are difficult to determine globally. For these models, thermal remote sensing data (e.g., land surface temperature, LST) are the most important inputs. However, using the 8-day composite MODIS LST (the average LST of all cloud-free data in the compositing window) (Wan et al., 2002) and daily meteorological data recorded at the flux tower, Cleugh et al. (2007) demonstrate that the results from thermal models are unreliable at two Australian sites (Virginia Park, a wet/dry tropical savanna located in northern Queensland and Tumbarumba, a cool temperate, broadleaved forest in south east New South Wales). Using a combination of remote sensing and global meteorological data, we have adapted the Cleugh et al. (2007) algorithm, which is based on the Penman–Monteith method and calculates both canopy conductance and ET. Monteith (1965) eliminated surface temperature from Equations (1) – (3) to give:
where , the slope of the curve relating saturated water vapor pressure (esat) to temperature; A′ is available energy partitioned between sensible heat and latent heat fluxes on land surface. VPD = esat – e is the air vapor pressure deficit. All inputs have been previously defined except for surface resistance rs, which is an effective resistance accounting for evaporation from the soil surface and transpiration from the plant canopy.
Despite its theoretical appeal, the routine implementation of the P-M equation is often hindered by requiring meteorological forcing data (A', Ta and VPD) and the aerodynamic and surface resistances (ra and rs). Radiation and soil heat flux measurements are needed to determine A′; air temperature and humidity to calculate VPD; and wind speed and surface roughness parameters to determine ra. Multi-temporal implementation of the P-M model at regional scales requires routine surface meteorological observations of air temperature, humidity, solar radiation and wind speed. Models for estimating maximum stomatal conductance including the effect of limited soil water availability and stomatal physiology requires either a fully coupled biophysical model such as that by Tuzet et al. (2003) or resorting to the empirical discount functions of Jarvis (1976), which must be calibrated. Determining a surface resistance for partial canopy cover is even more challenging with various dual source models proposed (e.g. Shuttleworth and Wallace, 1985) to account for the presence of plants and soil.
联系客服
