Open-Earth-Monitor Global Workshop 2025

Abstract
We monitor the impact of 320 nature-based climate solutions (NBS) implemented through carbon offset projects across 55 countries using a standardized methodology based on Earth Observations (EO) Big Data. Our objective is to demonstrate the feasibility for using free and open EO data at high and low moderate spatial resolutions to support M&E. We identify current gaps and provide recommendations both for technical enhancements and for the design of restoration policies. Finally, we deploy a Google Earth Engine app called ‘’Monitor-EO’’, which allows the end user to perform the analysis described in the paper, using a graphic user interface (GUI).
We employ an ecosystem-based approach to assess impacts, by simultaneously monitoring three key ecosystem variables such as vegetation cover, land surface temperature and soil moisture. We assume that a positive outcome from a restoration measure would lead to increased vegetation levels, increased soil moisture, and decreased temperature of the soil surface.
Comparison areas are defined for each project restoration site using a 2km buffer around each restoration area and one additional comparison area which i) is randomly created within a radius of 3 kilometres from the restoration area, and ii) has equal area than the restoration site?.
We measure across restoration and control sites three variables, the Normalized Difference Vegetation Index (NDVI), the Normalized Difference Water Index (NDWI), and the Land Surface Temperature (LST), three years before the restoration activity was implement and then for the years following until 2024. We use different EO products from MODIS data (NDVI and NDWI at 500 meters spatial resolution and the LST at 1000 meters spatial resolution)
For each variable we measure the Difference-in-Difference (DiD), and we perform a trend analysis. We run a series of statistical tests to ascertain the statistical significance of changes in order to infer causality of the project interventions. Finally, for selected project sites, we perform a Spatial Autocorrelation Analysis to assess the degree of spatial clustering of positive changes (indicative of restoration success) compared to a random distribution.

The Monitor-EO application identified significant trends in at least one environmental indicator (NDVI, LST, or NDWI) in 62% (199) of the 320 projects analysed, covering all regions except Europe. Globally, NDVI exhibited predominantly positive trends, with 92% of projects showing increases, particularly in North America, Asia, Latin America and the Caribbean (LAC), and Africa, indicating favourable environmental changes. In contrast, LST displayed decreasing trends in 54% of the projects, with the most pronounced reductions observed in Africa and Asia. NDWI, however, exhibited declining trends in the majority of projects, with only 19% showing increases, primarily in Africa and North America. Projects demonstrating the highest rates of change were initiated in 2011 and are projected to extend for over 20 years. Smaller projects (less than 1,000 hectares) exhibited more pronounced trends compared to larger projects, while longer monitoring periods (exceeding 10 years) were associated with more substantial and statistically significant trends.

Keywords: earth observations, monitoring and evaluation, landscape restoration, ecosystem health, climate change adaptation, food security,