Abstract
Satellite-based data from the MODIS sensor, such as temperature or vegetation indices, are increasingly being used in landscape epidemiology research (especially in data-poor areas), e.g. to model the spatio-temporal distribution of vectors or to better understand their bio-ecology. However, due to their size and spatio-temporal granularity, these data can be difficult to access and process, especially when large time series are required. Here we present modis2fast, an R package that provides a simple and efficient way to download MODIS data (as well as VIIRS and GPM data) in R. Built upon the robust open-source OPeNDAP framework, modis2fast allows the user to subset any MODIS product directly at the download stage, using spatial, temporal, and band/layer filters ; thus reducing the downloaded data and disk usage to the strictly necessary. In addition, modis2fast supports parallel downloads. This package is therefore particularly useful for retrieving and processing large MODIS time series at their finest spatial and temporal resolution. In this poster, we also present some practical examples of the use of modis2fast for research on mosquito-borne diseases. modis2fast naturally promotes digital sobriety in our research. The package is available at https://github.com/ptaconet/modis2fast.