Scholarlink

<p>Accurate data on rice distribution is crucial for cultivation management in Bangladesh. Remote sensing provides a complementary solution to traditional methods. Prior studies often used fixed thresholds for rice area estimation, neglecting the impact of threshold variation on mapping accuracy. This study addresses this gap by evaluating thresholds (T = 0.00–0.07) to optimize Boro rice mapping in Bangladesh (2011–2017) using Moderate Resolution Imaging Spectroradiometer (MODIS)-derived Enhanced Vegetation Index (EVI) and Land Surface Water Index (LSWI) at 500-meter resolution. The temporal relationship between EVI and LSWI was analyzed pixel-by-pixel during the Boro growing season (late November to early July), focusing on cropland areas. Rice fields were considered when LSWI exceeded EVI at least once during transplanting. Spectral indices were extracted from 178 rice fields across 33 districts to examine the EVI-LSWI relationship. MODIS-derived rice areas were validated against national census data from the Bangladesh Bureau of Statistics (BBS). Results indicated that LSWI consistently exceeded EVI during transplanting in most districts, supporting its reliability for large-scale rice mapping. A no-threshold condition (T = 0.00) provided the most accurate estimates, with moderate bias (-67.7&#xa0;km²), average RMSE (293.3&#xa0;km²; 0.6%), and strong performance (NSE = 0.73). Misalignments were notable in flood-prone, coastal, and physio-graphically distinct districts. This method effectively captures the seasonal dynamics of rice cultivation with improved accuracy. Its simplicity and reduced reliance on ground-truth data make it suitable for resource-limited regions. The study demonstrates that this strategy offers a scalable, reliable tool for enhancing rice monitoring and management practices.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Mapping Boro Rice Cultivation in Bangladesh Using Multi-Temporal MODIS Data and Phenological Approach

  • 期刊论文

  • 发表日期：

    2025年2月7日

查看全文
  • Md. Mizanur Rahman,
  • Nitin Kumar Tripathi,
  • Chitrini Mozumder,
  • Siwat Kongwarakom,
  • Salvatore Gonario Pasquale Virdis

摘要

Accurate data on rice distribution is crucial for cultivation management in Bangladesh. Remote sensing provides a complementary solution to traditional methods. Prior studies often used fixed thresholds for rice area estimation, neglecting the impact of threshold variation on mapping accuracy. This study addresses this gap by evaluating thresholds (T = 0.00–0.07) to optimize Boro rice mapping in Bangladesh (2011–2017) using Moderate Resolution Imaging Spectroradiometer (MODIS)-derived Enhanced Vegetation Index (EVI) and Land Surface Water Index (LSWI) at 500-meter resolution. The temporal relationship between EVI and LSWI was analyzed pixel-by-pixel during the Boro growing season (late November to early July), focusing on cropland areas. Rice fields were considered when LSWI exceeded EVI at least once during transplanting. Spectral indices were extracted from 178 rice fields across 33 districts to examine the EVI-LSWI relationship. MODIS-derived rice areas were validated against national census data from the Bangladesh Bureau of Statistics (BBS). Results indicated that LSWI consistently exceeded EVI during transplanting in most districts, supporting its reliability for large-scale rice mapping. A no-threshold condition (T = 0.00) provided the most accurate estimates, with moderate bias (-67.7 km²), average RMSE (293.3 km²; 0.6%), and strong performance (NSE = 0.73). Misalignments were notable in flood-prone, coastal, and physio-graphically distinct districts. This method effectively captures the seasonal dynamics of rice cultivation with improved accuracy. Its simplicity and reduced reliance on ground-truth data make it suitable for resource-limited regions. The study demonstrates that this strategy offers a scalable, reliable tool for enhancing rice monitoring and management practices.