A simplified method to separate latent and sensible heat fluxes using remotely sensed data

In this paper, the relationships between NDVI (normalized difference vegetation index) and pseudo thermal inertia (land surface temperature difference between day and night AVHRR data) are discussed. A simplified method was established to separate the latent and sensible heat fluxes on vegetation covered land surface. Analogous correlations between surface reflectance and LST (Land Surface Temperature) had been illustrated by several scientists who are concerned about regional estimation of soil evaporation in and or semi-arid areas. It proves work well in retrieving of soil evaporation or even aerodynamic resistance of beat transfer. The scattergram of pseudo thermal inertia versus NDVI demonstrates that higher frequency points clearly form a triangle shape, that is to say, the higher frequency points are within the triangle. The upper side of the triangle can be written as: DT_u(NDVI)=a_u+b_u*NDVI. Similarly, the lower side of it can be denoted as: DT_l(NDVI)=a_l+b _l*NDVI. Then, Bowen ratio B_i can be expressed as following: B_i=(DT_u(NDVI_i)-DT_i)/(DT _u(NDVI_i)-DT_l(NDVI_i))-1, where DT_i is the land surface temperature difference between day and night of the ith pixel, NDVI_i is as well. The soil heat flux G is determined empirically. After separating Rn (net radiation) into Rn_s and Rn_v (net radiation for soil and vegetation respectively) based on the vegetation fraction, the G can be obtained by: G=0.35* Rn_s+0.05* Rn_v, where the constant parameters may vary according to different land cover and land use