VARIATIONS IN THE YIELD OF NATURAL STEPPE PASTURES IN KAZAKHSTAN WITH DIFFERENT HUMIDIFICATION BY SATELLITE DATA

Abstract

The brief review of current approaches of the seasonal productivity calculation in natural pastures by remotely sensed data is given. Results and comparison of actual and averaged productivity estimation are discussed for Naurzum Nature Reserve, chosen as the example of the area with minimal human impact. The sensitivity of some spectral indices to the characteristics of vegetation is discussed. Correlation between ground–measured productivity and spectral indices varies from high to very high. The SAVI is recommended to apply when the study area is represented with mixed landscapes and high content of intrazonal vegetation communities. SAVI is sensitive to chlorophyll concentration in green biomass, as well as it reduces the signal of bare soils. SAVI demonstrates better sensitivity to the subtle variations of vegetation conditions in mixed landscapes than other vegetation indices do. The equation, based on SAVI, was used to estimate and to compare seasonal pasture productivity of dry and wet years. It is concluded that the main loss of the productivity during dry year takes place in spring and early-summer. The article substantiates the feasibility and the need to introduce algorithms and the results of satellite data processing into the activities of Natural Reserves.