Agriculture Services

Detection of Flooded Rice Fields

This service detects stands that are under water in the rice growing area at the beginning of the agricultural season. The treatment is based on the combination of vegetation and moisture indices (NDVI and NDWI), as well as on the transformation of the image into Hue/Saturation/Value (HSV) allowing to highlight the water in blue. For each processed image, the extent of the areas covered by water is provided in raster format, allowing the evolution of the areas under water to be monitored over the season. If plot boundaries are provided by the user, the monitoring is then limited to the plots, thus avoiding the detection of water outside the rice growing area (pond formations, watercourses).

The images used to monitor the underwatering are Sentinel-2 images with a spatial resolution of 10 or 20 m depending on the spectral bands.

Ploughing Detection  

This service detects areas in an agricultural environment showing signs of ploughing. The treatment is based on the detection of change between two successive images during the ploughing period and detects areas that change from light brown (a sign of bare, dry soil) to dark brown (a sign of overturned, wetter soil). For each pair of images processed, the extent of the ploughed areas is provided in raster format, allowing the monitoring of ploughing changes over the course of a season. This service is relatively sensitive to cloud cover and shadows.

The images used to monitor the ploughing are Sentinel-2 images with a spatial resolution of 10 or 20 m depending on the spectral bands.

Estimated Growth Start Date   

This service illustrates the evolution of the vegetation cover of a series of agricultural plots within a graph. It calculates for each date when an image is available, the average of the pixels of the Sentinel-1 SAR images within each plot for which boundaries are to be provided. These values are recorded in a table allowing the evolution of the signal over a season to be observed. Generally speaking, in agricultural areas, the intensity of the pixels in an SAR image is correlated with the presence of vegetation.  Thus, when the ground is bare, the SAR intensity is assumed to be low, while as the vegetation develops, the SAR intensity increases. The beginning of the agricultural season is marked by the transition of the signal from low and stable values to values that increase over time. The service therefore proposes to estimate the date of this inflection point, which represents a key indicator for the course of the season.

The images used to estimate the date of the start of growth are Sentinel-1 images with a spatial resolution of 10 m.

Estimation of Phenological Parameters 

This service calculates the average of the Normalized Difference Vegetation Index (NDVI measures vegetation changes). The intensity of the NDVI gives an indication of the stage of cultivation since the stronger the NDVI, the denser and greener the vegetation cover. The phenology service thus proposes to estimate the date of three phenological indicators linked to the evolution of NDVI on a plot during a crop year. These indicators are 1) the beginning of growth, 2) the date of maximum growth, and 3) the date of senescence.

The images used to monitor the phenology are Sentinel-2 images with a spatial resolution of 10 or 20 m depending on the spectral bands.

Reference Projects

SPACEBEL is a key player in the NADiRA project, a Horizon 2020 innovation action that aims to develop agriculture in Africa using sustainable digital farming solutions. NADiRA integrates the industrialisation of Earth Observation products, data from connected sensors (IoT) and mobile devices which provide stakeholders with key information to invest in smallholder farming.