In the face of limitations in revisit times and frequent cloud coverage, many land and water applications need to combine data from multiple medium or high resolution sensors in order to build a time series.
In Dec 2017, the BELHARMONY research project started off with the aim to assess the consistency between recent, high resolution Earth Observation sensors, focusing on PROBA-V (100m), Sentinel-2 and Landsat-8 and DEIMOS-1. It addresses for instance biases between the sensors at early processing stages (level-1, top-of-atmosphere) using a series of calibration sites. An example is the dry lakebed of the Railroad Valley Playa calibration site in Nevada, USA, shown in this week’s 100m PROBA-V image.
In addition, BELHARMONY looks into the differences in the relative spectral response (for level-2 reflectance products) and the use of a common processing chain towards a set of derived land and water parameters (level-3 processing). To evaluate the impact of the harmonization measures, the project relies on EO and in situ data collection over the network of urban, agricultural and coastal sites from the Belgian BELAIR network.