Validating a Dynamic Global Vegetation Model with Remotely Sensed Vegetation Index

摘要

The present study aims to evaluate the ability of IBIS model to capture the difference in vegetation characteristics among six major biomes in the Northeast China Transect and to calibrate the simulated LAI by IBIS, using the product of MODIS LAI (Leaf Area Index). The results showed that IBIS simulated a little lower growing season LAI over temperate evergreen conifer forest and boreal evergreen forest, while it overestimated LAI relative to MODIS in non-growing season. IBIS performed poorly on LAI over savanna, grassland and shrub land, compared with MODIS and it nearly simulated higher LAI throughout the year. Based on regression analysis, the simulating LAI by IBIS (Integrated Biosphere Simulator) presented a significant linear correlation with that from MODIS over temperate evergreen conifer forest in spring and winter, boreal evergreen forest throughout the year and grassland from summer to early autumn. Therefore, it was help to adjust the model parameters over these plant functional types to calibrate the estimated LAI in a large spatial scale.