In-situ validation of Land Surface Phenology, Land Surface Temperature and Surface Water derived from Earth Observation products: Doñana protected area as a potential cal/val supersite

In this paper we assess spatial and temporal correlation of different available Earth Observation (EO) products provided through downstream services with in-situ measurements for validation purposes. We selected three relevant ecosystem indicators such as Land Surface Phenology, Land Surface Temperature and Surface Water, being widely available as standard remote sensing products using different methods and satellite sensors. The study seeks to contribute to develop a global framework to address the urgent need for coordinated cal/val activities of EO products. As a pilot, we made use of systematic data collected at Doñana LTSER Platform, one of the long-term measuring sites from the European Research Infrastructure eLTER. In order to enhance EO use we developed the GeeLTERMap tool and PyVPP python package to enable the retrieval of the selected EO products for any site. Phenology metrics estimated by in situ phenocams were used to assess Copernicus High Resolution Vegetation Phenology and Productivity (VPP), MODIS MCD12Q2 and PhenoPy metrics. In situ measurements by calibrated broadband Infrared Radiometers were used to validate LST retrieved from MODIS, Landsat and Sentinel-3 images. Finally, field sampling and in situ automatic camera photos were used to evaluate different multispectral indices and simple bands to map water occurrence using Sentinel-2 images. Results for phenology validation show HR-VPP as the most accurate EO product with an average Root Mean Square Error (RMSE) of 56 days for all metrics across different land covers. For LST, Landsat showed lower RMSE LST values than Sentinel-3 SLSTR and MODIS MOD11A1 products. Finally, Sentinel-2 MSI Band 12 was the most accurate band to delineate water bodies of Doñana shallow marshes. The developed tools and protocols for validation using long-term data from in situ sampling and measurements will be one of the services provided by eLTER Research Infrastructure.