Monitoring crop and natural vegetation conditions is highly relevant, particularly in the food insecure areas of the world. Data from remote sensing image time series at high temporal and medium to low spatial resolution can assist this monitoring as they provide key information about vegetation status in near real-time over large areas. The Software for the Processing and Interpretation of Remotely sensed Image Time Series (SPIRITS) is a stand-alone flexible analysis environment created to facilitate the processing and analysis of large image time series and ultimately for providing clear information about vegetation status in various graphical formats to crop production analysts and decision makers. In this paper we present the latest functional developments of SPIRITS and we illustrate recent applications. The main new developments include: HDF5 importer, Image re-projection, additional options for temporal Smoothing and Periodicity conversion, computation of a rainfall-based probability index (Standardized Precipitation Index) for drought detection and extension of the Graph composer functionalities. In particular,. The examples of operational analyses are taken from several recent agriculture and food security monitoring reports and bulletins. We conclude with considerations on future SPIRITS developments also in view of the data processing requirements imposed by the coming generation of remote sensing products at high spatial and temporal resolution, such as those provided by the Sentinel sensors of the European Copernicus programme.

This paper presents the results of a sub-pixel classification of crop types in Bulgaria from PROBA-V 100 m normalized difference vegetation index (NDVI) time series. Two sub-pixel classification methods, artificial neural network (ANN) and support vector regression (SVR) were used where the output was a set of area fraction images (AFIs) at 100 m resolution with pixels containing estimated area fractions of each class. High-resolution maps of two test sites derived from Sentinel-2 classifications were used to obtain training data for the sub-pixel classifications. The estimated area fractions have a good correspondence with the true area fractions when aggregated to regions of 10 × 10 km², especially when the SVR method was used. For the five dominant classes in the test sites the R² obtained after the aggregation was 86% (winter cereals), 81% (sunflower), 92% (broad-leaved forest), 89% (maize), and 67% (grasslands) when the SVR method was used.

Multitemporal Landsat7-ETM+ imagery and supplementary field survey data was used to establish a crop and land use map for an area of ±20 000 km² of Sanjiang Plain in China’s north-eastern province Heilongjiang. The “field-based” map is based on data of the year 2002 and is created by combining the results of a supervised Maximum Likelihood classification with the estimated field boundaries extracted by an automatic segmentation procedure. No sufficient ground data was available to perform an independent validation, and some classes are certainly uncertain. Nevertheless, given the minimal investments and the unexplored potentials for improvement, the approach seems very promising and cost-effective. By superimposing the classification with the map of the administrative counties, regional statistics were derived with the acreage distributions of the concerned crops and land use types.