Normal view MARC view ISBD view

Tanmay Mondal

Understanding the impact of urbanisation on wetlands using random forest machine learning algorithm in Google Earth Engine: a case of East Kolkata Wetlands, India - P 608-617

Land cover maps help in understanding the spatial and
temporal dynamics of a particular area in a given time.
The present study was done in the Kolkata Metropolitan Area (KMA) and East Kolkata Wetland (EKW) in
India using the machine learning (ML) algorithm of
random forest classification and prediction in Google
Earth Engine. The ML algorithm was used to classify
the land use land cover (LULC) maps of KMA and
EKW for 2002, 2013 and 2023. The ML prediction
model further used these LULC maps to predict the future land cover map for the year 2033, both for KMA
and EKW. A correlation analysis was done between
the built-up of KMA and the built-up of EKW to assess
the impact of the urban expansion of KMA on the wetlands of EKW. The present study explored the impact
of the increase in built-up area on the wetland and it
was seen that the built-up of EKW was increasing due
to the increase in the urban expansion of KMA. As a
result, the wetland is experiencing a decrease in biodiversity and quality of water and soil.

Subjects--Facted Index Terms:
Change detection
random forest
Google Earth Engine
machine learning
prediction