With the increasing availability of high-resolution and near real-time satellite observations, an opportunity arises to monitor and predict crop yield and drought stress impact on agricultural crop production on a pan-European scale. An international consortium of KnowH2O, Moisture Matters, Wageningen University & Research, Hochschule Rhein-Waal (D) and StellaSpark set out to develop a grassland monitoring service that does just that. The service, developed within the EU Interreg SPECTORS project, allows governments to estimate grassland yields for statistical and accounting purposes and track drought impact on individual parcel level.

The grassland monitoring service uses StellaSpark Nexus as platform to maintain a near real-time digital twin of an area covering the entire Netherlands and part of Germany, the state of Nordrhein-Westfalen. This digital twin is essentially a digital clone of all relevant objects and observations related to agriculture and water management, consisting of nearly 3.5 million agriculture parcels with their current crop types, close to 4,000 nature reserves, 1,200 weather stations and daily high-resolution multispectral satellite observations. Goal of the monitoring service is to provide up-to-date crop status information and to estimate optimal harvesting/cutting times for grassland, based on state-of-the-art data sources of ESA and official data sources of EU member states. Nexus has been connected to all these sources and provides standardisation of units, spatial reference systems and time zones, as well as visualization through an online dashboard.

Two specialized algorithms run on the digital twin; FluxPark, developed by Moisture Matters, which calculates soil moisture availability, actual crop transpiration, and estimated dry matter harvest yields on a daily basis. Its companion algorithm Sen2Grass, developed by Wageningen University & Research, estimates daily changes in biomass based on satellite observations. Algorithm performance has been validated by comparing model results for the years 2019 and 2020 with drone observations by Hochschule Rhein-Waal (D) and actual harvest data gathered at a test site of Haus Riswick in Kleve (D). The resulting information is available for individual parcels covering the entire application area.

Development of the grassland monitoring service was partly financed through the EU’s Interreg programme.

Near real-time grassland monitoring