Enhanced Polarimetric Radar Vegetation Index and Integration with Optical Index for Biomass Estimation in Grazing Lands Across the Contiguous United States

Grazing lands are crucial for agricultural productivity, ecological stability, and carbon sequestration, underscoring the importance of monitoring vegetation biomass for the effective management of these ecosystems. Remote sensing data, including optical vegetation indices (VIs) like the Normalized...

Full description

Saved in:
Bibliographic Details
Main Authors: Jisung Geba Chang, Simon Kraatz, Martha Anderson, Feng Gao
Format: Article
Language:English
Published: MDPI AG 2024-11-01
Series:Remote Sensing
Subjects:
Online Access:https://www.mdpi.com/2072-4292/16/23/4476
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850060158079598592
author Jisung Geba Chang
Simon Kraatz
Martha Anderson
Feng Gao
author_facet Jisung Geba Chang
Simon Kraatz
Martha Anderson
Feng Gao
author_sort Jisung Geba Chang
collection DOAJ
description Grazing lands are crucial for agricultural productivity, ecological stability, and carbon sequestration, underscoring the importance of monitoring vegetation biomass for the effective management of these ecosystems. Remote sensing data, including optical vegetation indices (VIs) like the Normalized Difference Vegetation Index (NDVI), are widely used to monitor vegetation dynamics due to their simplicity and high sensitivity. In contrast, radar-based VIs, such as the Polarimetric Radar Vegetation Index (PRVI), offer additional advantages, including all-weather imaging capabilities, a wider saturation range, and sensitivity to the vegetation structure information. This study introduces an enhanced form of the PRVI, termed the Normalized PRVI (NPRVI), which is calibrated to a 0 to 1 range, constraining the minimum value to reduce the background effects. The calibration and range factor were derived from statistical analysis of PRVI components across vegetated regions in the Contiguous United States (CONUS), using dual-polarization C-band Sentinel-1 and L-band ALOS-PALSAR data on the Google Earth Engine (GEE) platform. Machine learning models using NPRVI and NDVI demonstrated their complementarity with annual herbaceous biomass data from the Rangeland Analysis Platform. The results showed that the Random Forest Model outperformed the other machine learning models tested, achieving R<sup>2</sup> ≈ 0.51 and MAE ≈ 498 kg/ha (relative MAE ≈ 32.1%). Integrating NPRVI with NDVI improved biomass estimation accuracy by approximately 10% compared to using NDVI alone, highlighting the added value of incorporating radar-based vegetation indices. NPRVI may enhance the monitoring of grazing lands with relatively low biomass compared to other vegetation types, while also demonstrating applicability across a broad range of biomass levels and in diverse vegetation covers.
format Article
id doaj-art-a836f75d58e54a5c968d6f5df954d0f5
institution DOAJ
issn 2072-4292
language English
publishDate 2024-11-01
publisher MDPI AG
record_format Article
series Remote Sensing
spelling doaj-art-a836f75d58e54a5c968d6f5df954d0f52025-08-20T02:50:40ZengMDPI AGRemote Sensing2072-42922024-11-011623447610.3390/rs16234476Enhanced Polarimetric Radar Vegetation Index and Integration with Optical Index for Biomass Estimation in Grazing Lands Across the Contiguous United StatesJisung Geba Chang0Simon Kraatz1Martha Anderson2Feng Gao3Hydrology and Remote Sensing Laboratory, USDA Agricultural Research Service, Beltsville, MD 20705, USAHydrology and Remote Sensing Laboratory, USDA Agricultural Research Service, Beltsville, MD 20705, USAHydrology and Remote Sensing Laboratory, USDA Agricultural Research Service, Beltsville, MD 20705, USAHydrology and Remote Sensing Laboratory, USDA Agricultural Research Service, Beltsville, MD 20705, USAGrazing lands are crucial for agricultural productivity, ecological stability, and carbon sequestration, underscoring the importance of monitoring vegetation biomass for the effective management of these ecosystems. Remote sensing data, including optical vegetation indices (VIs) like the Normalized Difference Vegetation Index (NDVI), are widely used to monitor vegetation dynamics due to their simplicity and high sensitivity. In contrast, radar-based VIs, such as the Polarimetric Radar Vegetation Index (PRVI), offer additional advantages, including all-weather imaging capabilities, a wider saturation range, and sensitivity to the vegetation structure information. This study introduces an enhanced form of the PRVI, termed the Normalized PRVI (NPRVI), which is calibrated to a 0 to 1 range, constraining the minimum value to reduce the background effects. The calibration and range factor were derived from statistical analysis of PRVI components across vegetated regions in the Contiguous United States (CONUS), using dual-polarization C-band Sentinel-1 and L-band ALOS-PALSAR data on the Google Earth Engine (GEE) platform. Machine learning models using NPRVI and NDVI demonstrated their complementarity with annual herbaceous biomass data from the Rangeland Analysis Platform. The results showed that the Random Forest Model outperformed the other machine learning models tested, achieving R<sup>2</sup> ≈ 0.51 and MAE ≈ 498 kg/ha (relative MAE ≈ 32.1%). Integrating NPRVI with NDVI improved biomass estimation accuracy by approximately 10% compared to using NDVI alone, highlighting the added value of incorporating radar-based vegetation indices. NPRVI may enhance the monitoring of grazing lands with relatively low biomass compared to other vegetation types, while also demonstrating applicability across a broad range of biomass levels and in diverse vegetation covers.https://www.mdpi.com/2072-4292/16/23/4476biomassCONUSgrazing landspolarimetric radar vegetation indexPALSARSentinel-1
spellingShingle Jisung Geba Chang
Simon Kraatz
Martha Anderson
Feng Gao
Enhanced Polarimetric Radar Vegetation Index and Integration with Optical Index for Biomass Estimation in Grazing Lands Across the Contiguous United States
Remote Sensing
biomass
CONUS
grazing lands
polarimetric radar vegetation index
PALSAR
Sentinel-1
title Enhanced Polarimetric Radar Vegetation Index and Integration with Optical Index for Biomass Estimation in Grazing Lands Across the Contiguous United States
title_full Enhanced Polarimetric Radar Vegetation Index and Integration with Optical Index for Biomass Estimation in Grazing Lands Across the Contiguous United States
title_fullStr Enhanced Polarimetric Radar Vegetation Index and Integration with Optical Index for Biomass Estimation in Grazing Lands Across the Contiguous United States
title_full_unstemmed Enhanced Polarimetric Radar Vegetation Index and Integration with Optical Index for Biomass Estimation in Grazing Lands Across the Contiguous United States
title_short Enhanced Polarimetric Radar Vegetation Index and Integration with Optical Index for Biomass Estimation in Grazing Lands Across the Contiguous United States
title_sort enhanced polarimetric radar vegetation index and integration with optical index for biomass estimation in grazing lands across the contiguous united states
topic biomass
CONUS
grazing lands
polarimetric radar vegetation index
PALSAR
Sentinel-1
url https://www.mdpi.com/2072-4292/16/23/4476
work_keys_str_mv AT jisunggebachang enhancedpolarimetricradarvegetationindexandintegrationwithopticalindexforbiomassestimationingrazinglandsacrossthecontiguousunitedstates
AT simonkraatz enhancedpolarimetricradarvegetationindexandintegrationwithopticalindexforbiomassestimationingrazinglandsacrossthecontiguousunitedstates
AT marthaanderson enhancedpolarimetricradarvegetationindexandintegrationwithopticalindexforbiomassestimationingrazinglandsacrossthecontiguousunitedstates
AT fenggao enhancedpolarimetricradarvegetationindexandintegrationwithopticalindexforbiomassestimationingrazinglandsacrossthecontiguousunitedstates