MODIS time series contribution for the estimation of nutritional properties of alpine grassland

Luigi Ranghetti, Bruno Bassano, Giuseppe Bogliani, Alberto Palmonari, Andrea Formigoni, Laura Stendardi, Achaz von Hardenberg
2016 European Journal of Remote Sensing  
12 Despite the Normalised Difference Vegetation Index (NDVI) has been used to make predictions on 13 forage quality, its relationship with bromatological field data has not been widely tested. This 14 relationship was investigated in alpine grasslands of the Gran Paradiso National Park (Italian Alps). 15 Predictive models were built using remotely sensed derived variables (NDVI and phenological 16 information computed from MODIS) in combination with geo-morphometric data as predictors of 17
more » ... ured biomass, crude protein, fibre and fibre digestibility, obtained from 142 grass samples 18 collected within 19 experimental plots every two weeks during the whole 2012 growing season. The 19 models were both cross-validated and validated on an independent dataset (112 samples collected 20 during 2013). A good predictability ability was found for the estimation of most of the 21 bromatological measures, with a considerable relative importance of remotely sensed derived 22 predictors; instead, a direct use of NDVI values as a proxy of bromatological variables appeared not 23 to be supported. 24 . Comparison the accuracies of 432 different spectral indices for estimation of vegetation cover fraction in sparse vegetated areas.
doi:10.5721/eujrs20164936 fatcat:uwo73u2jabf6ll3rblf573f6di