EU Science Hub

Development of a national and sub-national crop calendars data set compatible with remote sensing derived land surface phenology

Abstract: 
Crop calendars are a fundamental component of agricultural production monitoring systems since they help analysts to focus on the seasons when different crop types are actually growing in the field. The Earth Observation based early warning system ASAP (Anomaly hot Spots of Agricultural Production) uses land surface phenology (LSP) metrics as proxy for crop calendars and applies parameters, such as the start and end of the season (SOS and EOS respectively) to define the period of active agricultural vegetation growth at pixel level. However, such information is not crop specific and it remains therefore relevant to use crop calendars from independent sources providing crop specific key phenological timings, such as sowing, growing and harvesting. Several institutions, including FAO and USDA make available crop calendars at the national level, which are widely used for agricultural monitoring. The LSP derived SOS and EOS metrics can be associated with sowing and harvesting from such crop calendars. This report describes a method for the attribution of each growing season derived from LSP to a crop type listed in existing crop calendars. Based on a set of rules, we compare the growing seasons derived from LSP with the timings of the crop calendars, and select those crops where a match between LSP and crop calendar information is found. Agricultural statistics, including harvested area and production, are used in order to verify the correct identification and relevance of crop types by country. The method also allows to downscale the existing national level crop calendars to the sub-national level. It therefore makes available sub-national level crop calendars, which are highly valuable for crop monitoring at that scale. The resulting crop calendars are available in the ASAP download section: https://mars.jrc.ec.europa.eu/asap/download.php
JRC wide hidden block