Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Pays Bas → 2021 → Assessing and analysing desertification in the Chambal River Basin (CRB) using remote sensing and machine learning

UNESCO-IHE Institute for Water Education, Delft (2021)

Assessing and analysing desertification in the Chambal River Basin (CRB) using remote sensing and machine learning

Daiman, Amit

Titre : Assessing and analysing desertification in the Chambal River Basin (CRB) using remote sensing and machine learning

Auteur : Daiman, Amit

Université de soutenance : UNESCO-IHE Institute for Water Education, Delft

Grade : Master of Science (MS) 2021

Résumé partiel
Most research on the assessment of the desertification process is based on historical in-situ observations. There is a need to use large data for desertification mapping and monitoring to understand better the processes involved, make comprehensive planning, and prepare an action plan. The study aims to analyse and assess the trend and status of the desertification and impact of anthropogenic and meteorological drivers in Chambal River Basin (CRB), India, from 1990 to 2020 based on Machine Learning (ML) algorithm and remote sensing data. The Land Use and Land Cover map was prepared to analyse change detection to understand anthropogenic drivers’ impact on desertification. The LULC was prepared on a cloud computing-based platform called Google Earth Engine (GEE) using the Landsat series of satellite data. To get more accuracy and find the optimum ML algorithm for LULC mapping, the Random Forest (RF), Classification And Regression Tree (CART), and Support Vector Machine (SVM) supervised classification algorithm was applied. Kappa and overall accuracy were calculated for assessing the results, which showed more than 85% of accuracy in classification. RF supervised classification algorithm provided the highest accuracy for each LULC product from 1990 to 2020. The meteorological drought is an indicator to evaluate desertification in arid and semi-arid regions. To analyse the meteorological drought, the Global Land Data Assimilation System (GLDAS) Noah Land Surface Model (NLSM) soil moisture, rainfall products provided by the Indian Meteorological Department (IMD) and Standardised Precipitation-Evapotranspiration Index (SPEI) product by the Laboratory of Climate Service and Climatology (LCSC) from 1990 to 2020 were used. The SPEI based on different time scales (3, 6, 9, and 12 months) was used to evaluate the capability to assess the drought conditions to find the desertification process by a single index over CRB. SPEI was evaluated on the basin and district level.

Sujets  : desertification Chambal River Basin remote sensing machine learning cloud computing soil moisture land use land cover meteorological data

Présentation

Page publiée le 8 novembre 2022