A method for estimating the gross primary production of alpine meadows using MODIS and climate data in China

The use of remotely sensed data to estimate and monitor the gross primary production (GPP) of an ecosystem on regional scales is an important method in climate change research. Under the unremitting efforts of scientists, many successful remote-sensing-based GPP models have been developed for various vegetation types and regions. However, in practice, some models have been applied to a wide variety of ecosystems, and the suitability of a particular model for the environment under consideration has seldom been taken into account. Due to ecosystem diversity and climatic and environmental variation, it is often difficult to find a model that is suitable for a specific vegetation region. In this article, a new method is proposed for estimating the GPP of alpine vegetation, known as the alpine vegetation model (AVM). The accuracy of the AVM in estimating the GPP was compared to that of four other models: the vegetation photosynthesis model (VPM), eddy covariance-light use efficiency (EC-LUE) model, temperature and greenness (TG) model, and vegetation index (VI) model. The results demonstrated that the AVM displays superior accuracy in estimating the GPP of alpine vegetation. We also found that there is information redundancy in the input variables of these four models, which may account for their lower accuracy in estimating the GPP. In addition, the GPP estimates using the enhanced vegetation index are affected more in the case of low rather than high GPP by the influence of senesced grass during the early and late grassland growing season.