Louis Giglio , Luigi Boschetti , David P. Roy , Joanne V. Hall , Maria Zubkova , Michael Humber , Haiyan Huang , Vladyslav Oles
{"title":"The NASA VIIRS burned area product, global validation, and intercomparison with the NASA MODIS burned area product","authors":"Louis Giglio , Luigi Boschetti , David P. Roy , Joanne V. Hall , Maria Zubkova , Michael Humber , Haiyan Huang , Vladyslav Oles","doi":"10.1016/j.rse.2025.115006","DOIUrl":null,"url":null,"abstract":"<div><div>The VIIRS was designed in recognition of the need for long-term observations in support of global-change science and to provide MODIS observation continuity. The NASA MODIS burned area (BA) product (MCD64A1) has been generated systematically for nearly 25 years and maps globally the approximate day of burning at 500-m resolution. We describe the follow-on NASA Collection 2 VIIRS BA product (VNP64A1) that became available in October 2024 and provides global BA mapping from 2012 into the next decade. The product is derived using an adaptation of the Collection 6.1 MCD64A1 BA mapping algorithm and using the same sinusoidal grid and 500-m reporting resolution. Annual and monthly intercomparisons of the VPN64A1 and MCD64A1 products at global and continental/regional scales indicate high product BA reporting consistency. Approximately 2.6 % of the land surface was mapped as burned annually; VNP64A1 mapped slightly more than MCD64A1 but with 0-day median burn-date reporting differences. A global (Stage-3) product validation, using 561 interpreted Landsat image pairs, was undertaken and the VNP64A1 product had estimated 36.7 % commission and 72.4 % omission errors, compared to MCD64A1 34.2 % commission and 71.5 % omission errors. Product accuracies were also characterized using metrics derived from the weighted regression (to incorporate sample inclusion probabilities) between 5-km BA proportions in the products and the Landsat data. The omission and commission errors observed at 30-m were largely compensated at 5-km, with coefficient of determination (0.70), slope (0.83), and intercept (−0.0006) terms for the VNP64A1 product nearly identical to those of MCD64A1. With respect to global temporal reporting accuracy, 38 % of VNP64A1 burned grid cells had burn dates that matched the date of a MODIS active fire, versus 46 % for MCD64A1, and 71 % of VNP64A1 burned grid cells had burn dates within two days (72 % for MCD64A1). The research presented in this paper indicates that the VNP64A1 product is suitable for providing continuity of the MODIS BA record.</div></div>","PeriodicalId":417,"journal":{"name":"Remote Sensing of Environment","volume":"331 ","pages":"Article 115006"},"PeriodicalIF":11.4000,"publicationDate":"2025-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Remote Sensing of Environment","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0034425725004109","RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENVIRONMENTAL SCIENCES","Score":null,"Total":0}
引用次数: 0
Abstract
The VIIRS was designed in recognition of the need for long-term observations in support of global-change science and to provide MODIS observation continuity. The NASA MODIS burned area (BA) product (MCD64A1) has been generated systematically for nearly 25 years and maps globally the approximate day of burning at 500-m resolution. We describe the follow-on NASA Collection 2 VIIRS BA product (VNP64A1) that became available in October 2024 and provides global BA mapping from 2012 into the next decade. The product is derived using an adaptation of the Collection 6.1 MCD64A1 BA mapping algorithm and using the same sinusoidal grid and 500-m reporting resolution. Annual and monthly intercomparisons of the VPN64A1 and MCD64A1 products at global and continental/regional scales indicate high product BA reporting consistency. Approximately 2.6 % of the land surface was mapped as burned annually; VNP64A1 mapped slightly more than MCD64A1 but with 0-day median burn-date reporting differences. A global (Stage-3) product validation, using 561 interpreted Landsat image pairs, was undertaken and the VNP64A1 product had estimated 36.7 % commission and 72.4 % omission errors, compared to MCD64A1 34.2 % commission and 71.5 % omission errors. Product accuracies were also characterized using metrics derived from the weighted regression (to incorporate sample inclusion probabilities) between 5-km BA proportions in the products and the Landsat data. The omission and commission errors observed at 30-m were largely compensated at 5-km, with coefficient of determination (0.70), slope (0.83), and intercept (−0.0006) terms for the VNP64A1 product nearly identical to those of MCD64A1. With respect to global temporal reporting accuracy, 38 % of VNP64A1 burned grid cells had burn dates that matched the date of a MODIS active fire, versus 46 % for MCD64A1, and 71 % of VNP64A1 burned grid cells had burn dates within two days (72 % for MCD64A1). The research presented in this paper indicates that the VNP64A1 product is suitable for providing continuity of the MODIS BA record.
期刊介绍:
Remote Sensing of Environment (RSE) serves the Earth observation community by disseminating results on the theory, science, applications, and technology that contribute to advancing the field of remote sensing. With a thoroughly interdisciplinary approach, RSE encompasses terrestrial, oceanic, and atmospheric sensing.
The journal emphasizes biophysical and quantitative approaches to remote sensing at local to global scales, covering a diverse range of applications and techniques.
RSE serves as a vital platform for the exchange of knowledge and advancements in the dynamic field of remote sensing.