日本地球惑星科学連合2023年大会

講演情報

[E] オンラインポスター発表

セッション記号 M (領域外・複数領域) » M-AG 応用地球科学

[M-AG33] Satellite Land Physical Processes Monitoring at Medium/High/Very High Resolution

2023年5月24日(水) 15:30 〜 17:00 オンラインポスターZoom会場 (5) (オンラインポスター)

コンビーナ:Jean-Claude Roger(University of Maryland College Park)、Shinichi Sobue(Japan Aerospace Exploration Agency)、Eric Vermote(NASA Goddard Space Flight Center)、Ferran Gascon(European Space Agency)

現地ポスター発表開催日時 (2023/5/24 17:15-18:45)

15:30 〜 17:00

[MAG33-P02] Monitoring coffee lands through integration of DEM and Landsat data in Central Highlands, Vietnam

*Son Thanh Nguyen1、Chi-Farn Chen1Cheng-Ru Chen1、Youg-Sin Cheng1、Shih-Hsiang Chen2 (1.Center for Space and Remote Sensing Research, National Central University, Taiwan、2.Department of Finance and Cooperative Management, National Taipei University, Taiwan)

キーワード:Landsat, DEM, Random forests, Coffee, Change detection, Highlands Vietnam

Coffee is among the most valuable tropical export crops and the world’s most traded agricultural commodities. Vietnam is globally the second largest coffee exporter after Brazil. Coffee is the nation’s second most valuable agricultural export, playing considerable socioeconomic importance to provide income sources and rural livelihoods for millions of people living in the Central Highlands. Information on coffee-growing areas is thus important for policymakers to timely estimate the production to devise successful coffee-export strategies. The main objective of this study is to develop an approach for monitoring coffee-growing areas using the moderate-resolution Landsat satellite imagery and digital elevation model (DEM). The data were processed for 1995 and 2020 using the random forest (RF) algorithm, following three main steps: (1) data preprocessing to perform geometric corrections and derive vegetation indices, land surface temperature (LST), and topographic indices; (2) predictor importance analysis to determine predictor variables important for the model; (3) image classification using RF-based regression algorithm; and (4) accuracy assessment. The research findings indicated that 10 out of 43 variables (derived from satellite data and DEM) were critically important for coffee mapping. The classification results, derived from the model using such variables, compared with the ground reference data showed that the overall accuracies and Kappa coefficients were generally higher than 88.5% and 0.74, respectively. Such findings were reaffirmed by a close agreement with the official statistics, with the root mean squared error (RMSE) and mean absolute error (MAPE) values lower than 9.9% and 7.6%, respectively. From 1995 to 2020, the newly cultivated coffee area increased by 135,520 ha, due to the conversion of forests to coffee plantations. Such distributions of coffee areas and decadal changes in farming activities are critically useful for policymakers to formulate successful crop planning strategies. The findings obtained from this work in the form of quantitative information on spatiotemporal distributions and decadal changes in coffee-growing areas could be beneficial to land-use planners and agronomic policymakers to form successful crop management and agricultural planning strategies in the region.