Monitoring land cover and land use change is important for land resource planning, understanding ecosystem services including resilience to climate change, biodiversity conservation, decision making, and other purposes. A 2015 geospatial needs assessment of the Lower Mekong Region (Cambodia, Laos PDR, Myanmar, Thailand, and Vietnam) conducted by SERVIR-Mekong, a joint partnership with United States Agency for International Development (USAID) and the National Aeronautics and Space Administration (NASA), revealed a significant need for a comprehensive high-resolution land cover and land use classification mapping system for the region. Based on this needs assessment, various stakeholders from multiple agencies across the region came together to build Regional Land Cover Monitoring System (RLCMS) to improve regional land cover and land use mapping capacity in Lower Mekong. As a part of their involvement with SERVIR-Mekong project, Geospatial Analysis Laboratory (GsAL) at the University of San Francisco, co-organized a three day production workshop in Bangkok from January 17th-19th of 2017. This was the third RLCMS workshop that has brought together end users including representatives from government and non-government organizations, and educational institutes from across the region to collaboratively build a RLCMS that meets both regional and individual country and agency needs . Throughout the three-day workshop, end users assessed and critiqued the current land cover typology classes based on their specific needs, assembled map algorithms to output target land cover classes, evaluated accuracy assessment procedures, identified specific need based uses and shared sources for reference data collection. One of the key innovations of this build is the use of Google Earth Engine (GEE), a server based remote sensing platform for earth science data and analysis. It is through this platform that end users are able to see the results of the past two workshops come together as a system where they can now visualize land cover classes such as mangroves, leaf phenology and leaf type, impervious land and cropland among others.End users being regional experts helped in assessing accuracy of the primitives and shared ideas on improving the classification.
I was honored to represent the GsAL team at this conference. Over the last few months, I have been working collaboratively with the SERVIR-Mekong team to build out an algorithm for identifying mangroves using Google Earth Engine. Before working with the GsAL, I had never been exposed to the GEE platform. Using materials from the Google Earth Engine elective taught in the GsAL certificate program, she was able to understand the fundamentals of the programming language javascript and how it can be used to manipulate satellite data imagery. After getting accustomed with GEE, she began to work on building a script that used various bands and peer-reviewed reference data to identify mangroves in the Lower Mekong region over the years. The finalized script was then presented at the RLCMS workshop. Mangroves are endemic to Mekong Region and, thus, many of the participants were able to apply their regional knowledge to identify locations of mangroves and worked directly with me to improve the accuracy of her initial script. It is through collaborative engagements like this, USF and GsAL are contributing to research in the field of geospatial analysis and environmental management.
With the continued support from USF and end users in the region, RLCMS will be available in future with all the topological classifications for the Lower Mekong countries. As a part of this build, USF will be contributing to robusting policy decision, understanding ecosystem services, adapting to climate change resilience and supporting regional organization in the Lower Mekong. This collaboration will help in sustainable development of the region and lessons from this project could be translate for improving land use mapping all around the world.
I was honored to represent the GsAL team at this conference. Over the last few months, I have been working collaboratively with the SERVIR-Mekong team to build out an algorithm for identifying mangroves using Google Earth Engine. Before working with the GsAL, I had never been exposed to the GEE platform. Using materials from the Google Earth Engine elective taught in the GsAL certificate program, she was able to understand the fundamentals of the programming language javascript and how it can be used to manipulate satellite data imagery. After getting accustomed with GEE, she began to work on building a script that used various bands and peer-reviewed reference data to identify mangroves in the Lower Mekong region over the years. The finalized script was then presented at the RLCMS workshop. Mangroves are endemic to Mekong Region and, thus, many of the participants were able to apply their regional knowledge to identify locations of mangroves and worked directly with me to improve the accuracy of her initial script. It is through collaborative engagements like this, USF and GsAL are contributing to research in the field of geospatial analysis and environmental management.
With the continued support from USF and end users in the region, RLCMS will be available in future with all the topological classifications for the Lower Mekong countries. As a part of this build, USF will be contributing to robusting policy decision, understanding ecosystem services, adapting to climate change resilience and supporting regional organization in the Lower Mekong. This collaboration will help in sustainable development of the region and lessons from this project could be translate for improving land use mapping all around the world.