Regresar

Soil moisture content retrieval based on the MODIS for arid and semi-arid regions

Abstract:

In the arid and semi-arid regions, the vegetation's growing is controlled by the soil moisture. The soil moisture is an important indicator to show the water and energy exchanges between land surface temperature and atmospheric surface. This paper focuses on the modelling of thermal inertia and land surface temperature using MODIS data and the application study for the soil moisture and energy exchange, by taking Xinjiang Province as the representative region to calculate soil moisture content for 2005. In the model, the land surface temperature differences (△LST) and Broadband Albedo (α0) are used to assess the space-time variability of Soil Moisture Content (SMC). In order to the retrieval of soil moisture content for a 1 m depth soil profile from the ERS, a semi-empirical modelling approach is used. The model uses the simplified two layer water balance equation and the SMC is simulated basing on the linear relation with the Soil Moisture Saturation index (SMSI) at time t. For the validation, this paper uses the automatic weather gauging stations and the Neutron Probe monitoring actual SMC of 8 sites including Mosuowan, Tianshan Snow Site, Turpan Site, Bayangburk, Kuerle, Yinsu, Alagan, and Yiganbujima for 2005. When the time series of SMC-TI are compared with the records from the different sites, the slopes vary between 0.499 and 0.922. The R2 values are between 0.25 and 0.83 for the pooled case. The pooled RSME is 0. 015, the minimum and maximumRMSE is 0.001 and 0.028, respectively. So the apparent thermal inertia approach is suitable for the SMC retrieval in the arid and semi-arid region. The corresponding product of MODIS 10 day mean synthesis SMC can offer good quantitatively correct SMC images with the dramatic seasonal changes even without the observation sites.