Relating vegetation dynamics to temperature and precipitation at monthly and annual timescales in Taiwan using MODIS vegetation indices



To predict the responses of the timing, duration, and density of photosynthetically active plant cover to a changing climate, it is necessary to first quantitatively describe the relationships between temporal and spatial patterns of vegetation cover and both spatial and inter-annual variation in temperature and precipitation. We examined these relationships at multiple scales in Taiwan using monthly maximum composite values of MODIS-NDVI and MODIS-EVI between 2000 and 2012. The two vegetation indices were highly correlated to each other on a monthly basis for non-forest land-cover types, but correlated poorly in forests, probably due to the saturation of NDVI. However, the two indices were equally sensitive in detecting the onset and offset of growing season for all vegetation types. We found that EVI was positively related to both precipitation and temperature on a monthly timescale, although the relationships were not significant at the annual timescale. The much greater variation in monthly than in annual precipitation and temperature probably explains this difference. At low elevations, precipitation had a positive effect and temperature had a negative effect on EVI; however, at high elevations, which are mostly forested, both were positively related to EVI (although precipitation effects were not significant). We interpret this as evidence of water limitation of photosynthetic cover in the warmer, low-elevation parts of the island, whereas in the higher-elevation areas precipitation was usually adequate to satisfy evapotranspirative demand. This study illustrates the importance of examining the effects of precipitation and temperature on plant growth at a range of spatial and temporal scales. Specifically, finer spatial and temporal scales of analysis may better reveal climatic controls over vegetation growth than broader scales of analysis. ©


Earth Systems Research Center

Publication Date


Journal Title

International Journal of Remote Sensing


Taylor & Francis

Digital Object Identifier (DOI)


Document Type



2014 © Taylor & Francis.