[Study on the nutrition of alpine meadow based on hyperspectral data].

Remote sensing monitoring of alpine grassland nutritional status is a key factor of grassland reasonable utilization, also a difficulty for dynamic vegetation monitoring. The present paper studies the correlations between vegetation nutrition and hyperspectral data. The results showed that two band ratio models have a significant correlation with biomass, air-DM, P, CF, and CP. MAXR models have a significant correlation with most of nutrition index when selected wavebands equaled five. On the whole, the MAXR model precedes two band ratio models. Using MAXR models to estimate air-DM, P and CF can obtain higher accuracy.