Global land productivity datasets at 10-m spatial resolution

Remote sensing is the most cost-effective approach to assess global land change. Based on biophysical interactions between photosynthesis activity and narrow bands of the electromagnetic spectrum, vegetation indices have been widely used as proxy for land productivity and to monitor vegetation condition. Several global datasets representing land productivity are publicly available at no cost to the end user. However, these data normally feature coarse to medium spatial resolution; even when offered at medium scale, datasets have critical observation gaps given the revisiting time of the sensor combined with cloud coverage. Over the last decade the European Space Agency launched several new space-borne sensors under its Copernicus Programme, including two identical optical sensors Sentinel 2 MultiSpectral Instrument A & B - that collect images roughly every 5 days at 10-meter for the visual and near-infrared wavelengths. These relatively high spatial and temporal resolution Earth observation data combined with the computing processing power of cloud-based platforms have enabled the development of geospatial datasets representing biophysical variables covering the entire globe at fine-scale, offering a new perspective on how we monitor land from space. Here we used Google Earth Engine to apply a harmonic model to Sentinel-2A & Sentinel-2B imagery to derive annual integral land productivity at 10-m spatial resolution at global scale for 2018 to 2020 with minimal observation gaps. We developed global land productivity datasets based on Normalized Difference Vegetation Indices - NDVI and two alternative vegetation indices, the 2-band Enhanced Vegetation Index- EVI2, and the Modified Soil Adjusted Vegetation Index MSAVI. Although NDVI is the most commonly used vegetation index, EVI2 and MSAVI could provide improved sensitivity for measuring land productivity at locations with either very high or very low biomass. Global fine spatial resolution data on land productivity are critical for tracking changes in land condition from on-the-ground activities and to facilitate monitoring of different land management activities.