Open Access Open Access  Restricted Access Subscription Access

Deciphering Tropical Tree Communities Using Earth Observation Data and Machine Learning


Affiliations
1 Indian Institute of Remote Sensing (ISRO), Dehradun 248 001, India
2 National Remote Sensing Centre (ISRO), Hyderabad 500 037, India
 

Publicly available EO datasets offer new possibilities to generate biodiversity information at the community composition level, an essential biodiversity variable, beyond forest type. We demonstrated the potential of Sentinel-2, GEDI LiDAR canopy height and ALOSDEM in discriminating and classifying tropical tree communities in the Western Himalayas, India. For this, tree communities were first identified based on the ordination of field data and subsequently classified using satellite data applying machine learning, i.e. random forest (RF). From the three forest types in the study area, eight distinct tree communities were identified for which classification accuracy increased from single date (75.17%) to multi-date images (85.33%) and further by applying feature selection (88.17%). Whereas the best classification accuracy of 94.66% was achieved when canopy height and topographic variables were also considered. The findings suggest that RF is suitable for mapping tree communities by combining Sentinel-2 with GEDI and DEM parameters.

Keywords

Biodiversity, Canopy Height, Machine Learning, Remote Sensing, Tropical Forest.
User
Notifications
Font Size


  • Deciphering Tropical Tree Communities Using Earth Observation Data and Machine Learning

Abstract Views: 368  |  PDF Views: 203

Authors

Rahul Bodh
Indian Institute of Remote Sensing (ISRO), Dehradun 248 001, India
Hitendra Padalia
Indian Institute of Remote Sensing (ISRO), Dehradun 248 001, India
Divesh Pangtey
Indian Institute of Remote Sensing (ISRO), Dehradun 248 001, India
Ishwari Datt Rai
Indian Institute of Remote Sensing (ISRO), Dehradun 248 001, India
Subrata Nandy
Indian Institute of Remote Sensing (ISRO), Dehradun 248 001, India
C. Sudhakar Reddy
National Remote Sensing Centre (ISRO), Hyderabad 500 037, India

Abstract


Publicly available EO datasets offer new possibilities to generate biodiversity information at the community composition level, an essential biodiversity variable, beyond forest type. We demonstrated the potential of Sentinel-2, GEDI LiDAR canopy height and ALOSDEM in discriminating and classifying tropical tree communities in the Western Himalayas, India. For this, tree communities were first identified based on the ordination of field data and subsequently classified using satellite data applying machine learning, i.e. random forest (RF). From the three forest types in the study area, eight distinct tree communities were identified for which classification accuracy increased from single date (75.17%) to multi-date images (85.33%) and further by applying feature selection (88.17%). Whereas the best classification accuracy of 94.66% was achieved when canopy height and topographic variables were also considered. The findings suggest that RF is suitable for mapping tree communities by combining Sentinel-2 with GEDI and DEM parameters.

Keywords


Biodiversity, Canopy Height, Machine Learning, Remote Sensing, Tropical Forest.

References





DOI: https://doi.org/10.18520/cs%2Fv124%2Fi6%2F704-712