Optimal MODIS data processing for accurate multi-year paddy rice area mapping in China
Researchers, policy makers, and farmers currently rely on remote sensing technology to monitor crops. Although data processing methods can be different among different remote sensing methods, little work has been done on studying these differences. In order for potential users to have confidence in...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2020-07-01
|
| Series: | GIScience & Remote Sensing |
| Subjects: | |
| Online Access: | http://dx.doi.org/10.1080/15481603.2020.1773012 |
| _version_ | 1797679062887432192 |
|---|---|
| author | Li Liu Jingfeng Huang Qinxue Xiong Huijuan Zhang Peiling Song Yanhong Huang Yujie Dou Xiuzhen Wang |
| author_facet | Li Liu Jingfeng Huang Qinxue Xiong Huijuan Zhang Peiling Song Yanhong Huang Yujie Dou Xiuzhen Wang |
| author_sort | Li Liu |
| collection | DOAJ |
| description | Researchers, policy makers, and farmers currently rely on remote sensing technology to monitor crops. Although data processing methods can be different among different remote sensing methods, little work has been done on studying these differences. In order for potential users to have confidence in remote sensing products, an analysis of mapping accuracies and their associated uncertainties with different data processing methods is required. This study used the MOD09A1 and MYD09A1 products of the Moderate Resolution Imaging Spectroradiometer (MODIS) satellite, from which the Enhanced Vegetation Index (EVI) and the two-band EVI (EVI2) images were obtained. The objective of this study was to analyze the accuracy of different data processing combinations for multi-year rice area mapping. Sixteen combinations of EVI and EVI2 with two cloudy pixel removal methods (QA/BLUE) and four pixel replacement methods (MO/MY/MOY/MYO) were investigated over the Jiangsu Province of southeast China from 2006 to 2016. Different accuracy results were obtained with different data processing combinations for multi-year rice field mapping. Based on a comparison of the relative performance of different MODIS products and processing method combinations, EVI2_BLUE_MYO was proposed to be the optimal processing method, and was applied to forecasting the rice-planted area of 2017. Study results from 2006 to 2017 were validated against reference data and showed accuracies of rice area extraction of greater than 95%. The mean absolute error of transplanting, heading, and maturity dates were 11.55, 8.10, and 7.78 days, respectively. In 2017, two sample regions (A and B) were selected from places where rice fractional cover was greater than 75%. Rice area extraction accuracies of 85.0% (A) and 92.3% (B) were obtained. These results demonstrated the complementarity of MOD09A1 and MYD09A1 datasets in enhancing pixel spatial coverage and improving rice area mapping when atmospheric influences are significant. The optimal data processing combination indentified in this study is promising for accurate multi-year and large-area paddy rice information extraction and forecasting. |
| first_indexed | 2024-03-11T23:08:57Z |
| format | Article |
| id | doaj.art-f26c4d0aeffb4eb3aa754dc1e4885821 |
| institution | Directory Open Access Journal |
| issn | 1548-1603 1943-7226 |
| language | English |
| last_indexed | 2024-03-11T23:08:57Z |
| publishDate | 2020-07-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | GIScience & Remote Sensing |
| spelling | doaj.art-f26c4d0aeffb4eb3aa754dc1e48858212023-09-21T12:34:16ZengTaylor & Francis GroupGIScience & Remote Sensing1548-16031943-72262020-07-0157568770310.1080/15481603.2020.17730121773012Optimal MODIS data processing for accurate multi-year paddy rice area mapping in ChinaLi Liu0Jingfeng Huang1Qinxue Xiong2Huijuan Zhang3Peiling Song4Yanhong Huang5Yujie Dou6Xiuzhen Wang7Zhejiang UniversityZhejiang UniversityYangtze UniversityZhejiang UniversityZhejiang UniversityZhejiang UniversityZhejiang UniversityHangzhou Normal UniversityResearchers, policy makers, and farmers currently rely on remote sensing technology to monitor crops. Although data processing methods can be different among different remote sensing methods, little work has been done on studying these differences. In order for potential users to have confidence in remote sensing products, an analysis of mapping accuracies and their associated uncertainties with different data processing methods is required. This study used the MOD09A1 and MYD09A1 products of the Moderate Resolution Imaging Spectroradiometer (MODIS) satellite, from which the Enhanced Vegetation Index (EVI) and the two-band EVI (EVI2) images were obtained. The objective of this study was to analyze the accuracy of different data processing combinations for multi-year rice area mapping. Sixteen combinations of EVI and EVI2 with two cloudy pixel removal methods (QA/BLUE) and four pixel replacement methods (MO/MY/MOY/MYO) were investigated over the Jiangsu Province of southeast China from 2006 to 2016. Different accuracy results were obtained with different data processing combinations for multi-year rice field mapping. Based on a comparison of the relative performance of different MODIS products and processing method combinations, EVI2_BLUE_MYO was proposed to be the optimal processing method, and was applied to forecasting the rice-planted area of 2017. Study results from 2006 to 2017 were validated against reference data and showed accuracies of rice area extraction of greater than 95%. The mean absolute error of transplanting, heading, and maturity dates were 11.55, 8.10, and 7.78 days, respectively. In 2017, two sample regions (A and B) were selected from places where rice fractional cover was greater than 75%. Rice area extraction accuracies of 85.0% (A) and 92.3% (B) were obtained. These results demonstrated the complementarity of MOD09A1 and MYD09A1 datasets in enhancing pixel spatial coverage and improving rice area mapping when atmospheric influences are significant. The optimal data processing combination indentified in this study is promising for accurate multi-year and large-area paddy rice information extraction and forecasting.http://dx.doi.org/10.1080/15481603.2020.1773012modisdata processing combinationthe phenorice algorithmpaddy ricetime series |
| spellingShingle | Li Liu Jingfeng Huang Qinxue Xiong Huijuan Zhang Peiling Song Yanhong Huang Yujie Dou Xiuzhen Wang Optimal MODIS data processing for accurate multi-year paddy rice area mapping in China GIScience & Remote Sensing modis data processing combination the phenorice algorithm paddy rice time series |
| title | Optimal MODIS data processing for accurate multi-year paddy rice area mapping in China |
| title_full | Optimal MODIS data processing for accurate multi-year paddy rice area mapping in China |
| title_fullStr | Optimal MODIS data processing for accurate multi-year paddy rice area mapping in China |
| title_full_unstemmed | Optimal MODIS data processing for accurate multi-year paddy rice area mapping in China |
| title_short | Optimal MODIS data processing for accurate multi-year paddy rice area mapping in China |
| title_sort | optimal modis data processing for accurate multi year paddy rice area mapping in china |
| topic | modis data processing combination the phenorice algorithm paddy rice time series |
| url | http://dx.doi.org/10.1080/15481603.2020.1773012 |
| work_keys_str_mv | AT liliu optimalmodisdataprocessingforaccuratemultiyearpaddyriceareamappinginchina AT jingfenghuang optimalmodisdataprocessingforaccuratemultiyearpaddyriceareamappinginchina AT qinxuexiong optimalmodisdataprocessingforaccuratemultiyearpaddyriceareamappinginchina AT huijuanzhang optimalmodisdataprocessingforaccuratemultiyearpaddyriceareamappinginchina AT peilingsong optimalmodisdataprocessingforaccuratemultiyearpaddyriceareamappinginchina AT yanhonghuang optimalmodisdataprocessingforaccuratemultiyearpaddyriceareamappinginchina AT yujiedou optimalmodisdataprocessingforaccuratemultiyearpaddyriceareamappinginchina AT xiuzhenwang optimalmodisdataprocessingforaccuratemultiyearpaddyriceareamappinginchina |