Estimating sugarcane biomass is difficult to achieve when working with highly variable spatial distributions of growing conditions, like on Reunion Island. We used a dataset of in-farm fields with contrasted climatic conditions and farming practices to compare three methods of yield estimation based on remote sensing: (1) an empirical relationship method with a growing season-integrated Normalized Difference Vegetation Index NDVI, (2) the Kumar-Monteith efficiency model, and (3) a forced-coupling method with a sugarcane crop model (MOSICAS) and satellite-derived fraction of absorbed photosynthetically active radiation. These models were compared with the crop model alone and discussed to provide recommendations for a satellite-based system for the estimation of yield at the field scale. Results showed that the linear empirical model produced the best results (RMSE = 10.4 t∙ha⁻¹). Because this method is also the simplest to set up and requires less input data, it appears that it is the most suitable for performing operational estimations and forecasts of sugarcane yield at the field scale. The main limitation is the acquisition of a minimum of five satellite images. The upcoming open-access Sentinel-2 Earth observation system should overcome this limitation because it will provide 10-m resolution satellite images with a 5-day frequency.

This study explored the suitability of the Normalized Difference Vegetation Index (NDVI) from the Moderate Resolution Imaging Spectrometer (MODIS) obtained for six sugar management zones, over nine years (2002–2010), to forecast sugarcane yield on an annual and zonal base. To take into account the characteristics of the sugarcane crop management (15-month cycle for a ratoon, accompanied with continuous harvest in Western Kenya), the temporal series of NDVI was normalized through an original weighting method that considered the growth period of the sugarcane crop (wNDVI), and correlated it with historical yield datasets. Results when using wNDVI were consistent with historical yield and significant at P-value = 0.001, while results when using traditional annual NDVI integrated over the calendar year were not significant. This correlation between yield and wNDVI is mainly drawn by the spatial dimension of the data set (R² = 0.53, when all years are aggregated together), rather than by the temporal dimension of the data set (R² = 0.1, when all zones are aggregated). A test on 2012 yield estimation with this model realized a RMSE less than 5 t·ha⁻¹. Despite progress in the methodology through the weighted NDVI, and an extensive spatio-temporal analysis, this paper shows the difficulty in forecasting sugarcane yield on an annual base using current satellite low-resolution data. This is particularly true in the context of small scale farmers with fields measuring less than the size of MODIS 250 m pixel, and in the context of a 15-month crop cycle with no seasonal cropping calendar. Future satellite missions should permit monitoring of sugarcane yields using image resolutions that facilitate extraction of crop phenology from a group of individual plots.

Over the recent past, there has been a growing concern on the need for mapping cropping practices in order to improve decision-making in the agricultural sector. We developed an original method for mapping cropping practices: crop type and harvest mode, in a sugarcane landscape of western Kenya using remote sensing data. At local scale, a temporal series of 15-m resolution Landsat 8 images was obtained for Kibos sugar management zone over 20 dates (April 2013 to March 2014) to characterize cropping practices. To map the crop type and harvest mode we used ground survey and factory data over 1280 fields, digitized field boundaries, and spectral indices (the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Water Index (NDWI)) were computed for all Landsat images. The results showed NDVI classified crop type at 83.3% accuracy, while NDWI classified harvest mode at 90% accuracy. The crop map will inform better planning decisions for the sugar industry operations, while the harvest mode map will be used to plan for sensitizations forums on best management and environmental practices.

The objective of this study is to investigate the potential of TerraSAR-X (X-band) in monitoring sugarcane growth on Reunion Island (located in the Indian Ocean). Multi-temporal TerraSAR data acquired at various incidence angles (17°, 31°, 37°, 47°, 58°) and polarizations (HH, HV, VV) were analyzed in order to study the behaviour of SAR (synthetic aperture radar) signal as a function of sugarcane height and NDVI (Normalized Difference Vegetation Index). The potential of TerraSAR for mapping the sugarcane harvest was also studied. Radar signal increased quickly with crop height until a threshold height, which depended on polarization and incidence angle. Beyond this threshold, the signal increased only slightly, remained constant, or even decreased. The threshold height is slightly higher with cross polarization and higher incidence angles (47° in comparison with 17° and 31°). Results also showed that the co-polarizations channels (HH and VV) were well correlated. High correlation between SAR signal and NDVI calculated from SPOT-4/5 images was observed. TerraSAR data showed that after strong rains the soil contribution to the backscattering of sugarcane fields can be important for canes with heights of terminal visible dewlap (htvd) less than 50 cm (total cane heights around 155 cm). This increase in radar signal after strong rains could involve an ambiguity between young and mature canes. Indeed, the radar signal on TerraSAR images acquired in wet soil conditions could be of the same order for fields recently harvested and mature sugarcane fields, making difficult the detection of cuts. Finally, TerraSAR data at high spatial resolution were shown to be useful for monitoring sugarcane harvest when the fields are of small size or when the cut is spread out in time. The comparison between incidence angles of 17°, 37° and 58° shows that 37° is more suitable to monitor the sugarcane harvest. The cut is easily detectable on TerraSAR images for data acquired less than two or three months after the cut. The radar signal decreases about 5dB for images acquired some days after the cut and 3 dB for data acquired two month after the cut (VV-37°). The difference in radar signal becomes negligible (