Missing the Forest for the Trees: How accurate is the MODIS Vegetation Continuous Fields Tree Cover and does it matter?

Abstract:

The Earth observation product MODIS VCF has been widely used to estimate forest cover changes, parameterise vegetation and Earth System models, and as a reference for validation or calibration where field data is limited. However, although limited independent validations of MODIS VCF have shown that MODIS VCF’s accuracy decreases when estimating tree cover in sparsely-vegetated areas, such as in tropical savannas, no study has yet assessed the impact this may have on the VCF based tree cover distribution maps used by many in their research

Using tropical forest and savanna inventory data collected by the TROpical Biomes In Transition (TROBIT) project, we produce a series of corrections that take into account the spatial disparity between (i) the in-situ plot size and the MODIS VCF pixel, and (ii) the in-situ plot size and the Canopy Area Index (CAI). We then applied our corrections to areas identified as forest or savanna in the IGBP land cover mapping product. Our results show that the corrections are all weighted heavily by tree cover data of in-situ savanna plots, with all IGBP classes identified as ‘savanna’ showing substantial increases in cover after correction, indicating that MODIS VCF (Collection 6) consistently underestimates woody cover in tropical savannas. While more detailed in-situ field data is necessary to produce more accurate and reliable corrections, we recommend caution when using MODIS VCF (Collection 6) tree cover in tropical savannas, as the product estimates may be considerably lower than the actual tree cover. These underestimations in tropical savanna regions will have implications for research investigating forest-savanna tipping points and how forest-savanna distributions change in response to extreme weather events.