Hopp til hovedinnholdet

Publikasjoner

NIBIOs ansatte publiserer flere hundre vitenskapelige artikler og forskningsrapporter hvert år. Her finner du referanser og lenker til publikasjoner og andre forsknings- og formidlingsaktiviteter. Samlingen oppdateres løpende med både nytt og historisk materiale. For mer informasjon om NIBIOs publikasjoner, besøk NIBIOs bibliotek.

2025

Til dokument

Sammendrag

Mangrove forests are important ecosystems that store large amounts of carbon, protect coastlines, and provide habitats for a variety of marine species. Despite their ecological significance, mangrove forests are highly threatened. It is therefore crucial to map and monitor these areas to protect them. The purpose of this master's thesis is to map the distribution of mangrove forests and estimate their above-ground biomass in Ghana. Optical satellite images from Sentinel-2 were used for the classification. The biomass estimates were based on a combination of satellite images and LiDAR data from GEDI (Global Ecosystem Dynamics Investigation). In addition, fieldwork was conducted in Ghana to collect training and test data for the models. The global dataset Global Mangrove Watch (GMW) was used as a reference to assess the quality of the classification results. The distribution of the mangrove forest was mapped using Object-Based Image Analysis (OBIA), where the satellite images are divided into segments based on reflectance values in different spectral bands. The segments were then classified using the machine learning methods Support Vector Machine (SVM) and Random Forest (RF). To calculate the biomass, regression models were developed combining GEDI data with bands and indices from Sentinel-2. The models were evaluated based on the coefficient of determination (R²) and root mean square error (RMSE). The results show that the combination of Sentinel-2 data and local field measurements provides more accurate maps than existing global datasets. Among the two classification methods, SVM performed best on the validation data, with an F1-score of 96.5%, compared to 75.4% for GMW. The SVM model estimated the total area of mangrove forest in Ghana to be 143.54 km². The above-ground biomass in Ghana’s mangrove forest was estimated to be approximately 313,000 tons, with an average of about 21.8 tons per hectare. The best model for biomass estimation was the Gradient Boosting model, with a multiple regression analysis based on all attributes achieving an R² value of 0.38 and an RMSE of 24.5.

Til dokument

Sammendrag

Det er ikke registrert sammendrag