Cycle 7 (2018 Deadline)
Establishing a cropland database in Cambodia from remote sensing satellite data
PI: Sanara Hor (email@example.com), Royal University of Agriculture
U.S. Partner: Robert Hijmans, University of California, Davis
Dates: February 2019 - January 2021
Agriculture is a crucial sector in Cambodia, contributing more than one-quarter of the country’s gross domestic product and employing more than 8 million Cambodians. Having accurate information regarding land use in a user-friendly format would be a valuable tool for farmers and government planners alike. The overall purpose of this study is to establish cropland database in Cambodia using remote sensing satellite data. The specific objectives include (1) identifying appropriate satellite images and derived remote sensing products for cropland mapping; (2) developing a rapid cropland mapping tool based on cloud services for multiple time periods; (3) mapping different crop types across lowland and highland cropland regions; (4) exploring pathways to use the research results and technology to benefit Cambodian farmers; and (5) strengthening the research capacity of junior Cambodian researchers. Working in cooperation with their U.S. partners, the principal investigator and his team will examine remote sensing data from multiple sources including Landsat, MODIS, Sentinel-1, and Sentinel-2 and select the appropriate data for estimating cropland area in Cambodia based on availability, coverage, and cost. They will compile the data into a database that will cover specific sites representing complex cropping practices ranging from lowland to upland regions in the provinces located around Tonle Sap Lake. The team will present their results to various stakeholders through training programs, seminars, and lectures. The results and findings from the project will also be incorporated into the teaching curriculum at the Royal University of Agriculture.
In the past, cropland mapping has been unavailable to countries like Cambodia due to the cost. This project will provide an inexpensive means to map and monitor cropland from satellite images. The research will also provide basic cropland data that can be used in subsequent analyses like crop yield estimations or environmental impact assessments enabling better agricultural planning. Rapid crop mapping and measuring will also support the implementation of crop insurance products in Cambodia, which will be extremely beneficial for Cambodian farmers at risk of extreme weather events. In terms of facilitating better land use management, the project will also benefit government officials from various departments. The team will convey the results to them in an interactive way, using overlays in Google Maps, to assist in their planning efforts and encourage the diversification of agricultural land uses.
Summary of Recent Activities:
Beginning in August 2019, Dr. Hor and his colleagues began collecting aerial images of croplands in Battambang Province using a drone provided by their U.S. partners at the University of California, Davis. That process is complete as of March 2020, and team member Dr. Sophak has already processed the images, with RGB images being converted into a type of raster dataset called an orthophoto. This dataset is being used to guide ground truthing in a co-called Geopoint Data frame. The PI and his group will train this data for image classification with various models being employed for their general cropland database. In addition, the PEER team received free high-resolution satellite imagery from the USAID GeoCenter for image classification and analysis.
On January 9, 2020, Dr. Hor and his colleagues presented their technology to Minister Veng Sakhon of the Cambodian Ministry of Agriculture, Forestry, and Fisheries and U.S. Ambassador to Cambodia W. Patrick Murphy during their visit to the Royal University of Agriculture. The Minister encouraged the team to think broadly about big data science that can benefit the country, particularly with regard to agricultural development in Cambodia.
In late March 2020, the project encountered delays associated with the partial shutdown of the Royal University of Agriculture due to the COVID-19 crisis. However, staff continue to work from home, and in the coming months they will focus on completing their data analysis and then drafting the manuscript for a peer-reviewed paper. Later in 2020, they will organize a dissemination workshop or seminar to present the project’s results.
Back to PEER Cycle 7 Grant Recipients