An Optimal Temporal Resolution of Multispectral Satellite Data for Field-scale Agriculture

The objective of this study was to define the temporal resolution of future space missions dedicated to field-scale agriculture. The approach consisted in the assessment of the retrieval performance of the green area index (LAI) as a function of satellite revisit frequency. In this respect, different acquisition scenarios were simulated taking into consideration realistic LAI evolutions on large wheat fields, as well as realistic hypotheses: three levels of measurement errors and models uncertainties (10%, 20% and 25%), two probability levels of daily clouds occurrence (0.5 and 0.7), all these factors being combined with six scenarios of revisit frequency, covering the current capabilities of sensors (from 1 to 30 days). The high revisit frequency at high space resolution was routinely achieved by the concurrent use of three SPOT satellite sensors. The lack of continuous temporal coverage of estimated LAI values was overcome using a semi-empirical evolution model (MODLAI). The results of the simulations indicated that RMSE between the estimated and reference LAI values are low, and quite similar, up to 7 days revisit frequency, regardless of the error level or the cloudiness probability. These results were explained by the very good temporal interpolation performances of MODLAI, when the number of good quality acquisitions allowed its fine adjustment. * Corresponding author