Improving GPP estimates over cropland and grassland by using the Sentinel-2

Improving GPP estimates over cropland and grassland by using the Sentinel-2

21 Jun 2024

Monitoring the productivity over agricultural areas provides relevant information on the current vigorous growth and is critical input for yield estimation, and thus, food provision estimates. Traditionally, operational productivity datasets like the one from the Copernicus Global Land Monitoring Service (CGLMS) don’t provide field specific patterns due to the coarse spatial detail. Furthermore, current productivity algorithms are not specifically designed for agricultural land (cropland, grassland), which have specific dynamics that can’t be fully captured in the existing methods.

The EvoLand prototype aims to address these limitations and provide productivity estimates at field level detail over agricultural land (cropland & grassland), which is currently non-existing over Europe.

Sentinel-2 mission allows to upscale productivity monitoring to field level

The existing methods for estimating productivity and the related productivity indicators like Gross Primary Productivity (GPP) and Net Primary Productivity (NPP) generally rely on optical satellite data and meteorological data. From the optical data, a biophysical indicator can be derived showing the ratio of incoming photosynthetic active radiation that is absorbed by the plants for carbon assimilation. The launch of the optical Sentinel-2 satellite allows to upscale the productivity estimates to the detail of agricultural fields.

Different productivity indicators exist, which defines different stages of carbon storage. The most well‑known are the GPP and NPP. The GPP defines the total amount of carbon fixed during photosynthesis by the vegetation. From this carbon, approximately 50% will get immediately lost due to autotrophic respiration of the plant, the remainder of this is called the NPP. This defines the actual short‑term carbon storage of the plant and resulting biomass increase.

Evoland prototype provides GPP estimations specifically calibrated over agricultural land

The EvoLand prototype for GPP estimation offers new opportunities for productivity analysis by harnessing the capabilities of the Sentinel-2 satellite to achieve higher spatial detail in GPP estimation. This utilisation of Sentinel-2 optical satellite data together with meteorological data enables the EvoLand prototype to provide spatial detailed estimates of GPP in agricultural and grassland landscapes. Moreover, by calibrating the model for crop and grassland types, the prototype enhances the reliability of GPP estimates in these specific land cover categories.

One of the key features of the EvoLand prototype is its ability to offer 10-daily information on GPP over cropland and grassland locations, allowing for dynamic and detailed monitoring of productivity over time. This frequent temporal resolution provides valuable insights into the seasonal variations and trends in GPP, enabling users to make informed decisions regarding agricultural management practices. Furthermore, the integration of the Evoland prototype’s into OpenEO, enables a capacity to process data at the European level, which enhances its scalability and applicability across a broader geographical scope.

Guarantee consistency and quality of prototype

The EvoLand prototype undergoes rigorous consistency and quality checks to ensure the reliability and accuracy of its GPP estimation. The quality of the product is validated through comparisons with measurements from Integrated Carbon Observation System (ICOS) towers, which provide ground‑truth GPP data for calibration and validation purposes. This validation process helps to verify the accuracy of the GPP estimates generated by the EvoLand prototype.

To further validate the performance of the prototype, initial testing will be conducted in three distinct areas located in West Europe, Central Europe, and East Europe. By testing the prototype in geographically diverse regions, the robustness and generalisability of the GPP estimates can be assessed across different agro-ecological zones.

Additionally, the consistency of the current prototype is being verified by comparing its outputs with the Copernicus Global Land Monitoring service of GPP at a spatial resolution of 300 metres. This comparison should ensure that the GPP estimates produced by the EvoLand prototype still align with existing monitoring services.

This image presents an example of prototype output on GPP estimation over agricultural area in Belgium on first of August 2022.

Picture1

Continuous improvements of prototype during the Evoland project

During the EvoLand project the prototype will be further improved by elevating its capabilities in GPP estimation by incorporating novel enhancements. One of the key enhancements on the horizon is the assessment of integrating a drought stress efficiency term into the method. By incorporating this parameter, the EvoLand prototype aims to enhance its ability to better reflect the drought impact on GPP dynamics, which is currently lacking in the method and the existing operational CLMS.

Additionally, the upcoming improvements include temperature-specific parametrisation for crops, which will enable the prototype to differentiate GPP dynamics between these crop types based on different temperature requirements. This tailored parametrisation will enhance the accuracy of GPP estimates for C3 and C4 crops*, reflecting their distinct physiological responses to temperature variations.

Moreover, the EvoLand prototype is aiming to leverage other relevant environmental variables to further enhance the accuracy of GPP estimation. In this way, the prototype aims to improve the accuracy and reliability of GPP estimates, capturing the complex interplay of environmental factors on productivity.

Furthermore, the upcoming release of the HRL-VLCC products will define cropland and grassland locations on a yearly basis at European level, providing detailed information on crop type classes and grassland distribution. This information will enable the EvoLand prototype to focus specifically on these locations for GPP estimation.

* C3 plants are plants which utilise a metabolic pathway where the initial product of carbon dioxide assimilation is 3-phosphoglycerate.

C4 plants are a type of green plants that have adapted to regions with high sunlight, temperature, and limited water. They use a different CO2 fixation pathway compared to C3 plants, and have a higher photosynthetic capacity.

Sources: https://www.sciencedirect.com/topics/earth-and-planetary-sciences/c3-plant

https://www.sciencedirect.com/topics/earth-and-planetary-sciences/c4-plant

In-line image credit:  VITO

Cover image credit: European Union, Copernicus Sentinel-2 imagery

Would you be interested to follow the results of EvoLand and the development of this candidate prototype? Sign up for our newsletter (here below or via this form), follow us on LinkedInMastodon or X (Twitter). 

Questions about the article or the project? E-mail us at contact@evo-land.eu 

== 

This article is part of a series providing more details on all 11 EvoLand candidate prototypes.

Previous article in the series: 

Automated change detection with annual update for mapping urban dynamics – EvoLand (evo-land.eu)

 

Newsletter

Stay current! Subscribe to our EvoLand newsletter!