Exploring the potential of world view-2 red-edge band-based vegetation indices for estimation of mangrove leaf area index with machine learning algorithms

Yuanhui Zhu, Kai Liu, Lin Liu, Soe Myint, Shugong Wang, Hongxing Liu, Zhi He

Research output: Contribution to journalArticle

16 Citations (Scopus)

Abstract

To accurately estimate leaf area index (LAI) in mangrove areas, the selection of appropriate models and predictor variables is critical. However, there is a major challenge in quantifying and mapping LAI using multi-spectral sensors due to the saturation effects of traditional vegetation indices (VIs) for mangrove forests. WorldView-2 (WV2) imagery has proven to be effective to estimate LAI of grasslands and forests, but the sensitivity of its vegetation indices (VIs) has been uncertain for mangrove forests. Furthermore, the single model may exhibit certain randomness and instability in model calibration and estimation accuracy. Therefore, this study aims to explore the sensitivity of WV2 VIs for estimating mangrove LAI by comparing artificial neural network regression (ANNR), support vector regression (SVR) and random forest regression (RFR). The results suggest that the RFR algorithm yields the best results (RMSE = 0.45, 14.55% of the average LAI), followed by ANNR (RMSE = 0.49, 16.04% of the average LAI), and then SVR (RMSE = 0.51, 16.56% of the average LAI) algorithms using 5-fold cross validation (CV) using all VIs. Quantification of the variable importance shows that the VIs derived from the red-edge band consistently remain the most important contributor to LAI estimation. When the red-edge band-derived VIs are removed from the models, estimation accuracies measured in relative RMSE (RMSEr) decrease by 3.79%, 2.70% and 4.47% for ANNR, SVR and RFR models respectively. VIs derived from red-edge band also yield better accuracy compared with other traditional bands of WV2, such as near-infrared-1 and near-infrared-2 band. Furthermore, the estimated LAI values vary significantly across different mangrove species. The study demonstrates the utility of VIs of WV2 imagery and the selected machine-learning algorithms in developing LAI models in mangrove forests. The results indicate that the red-edge band of WV2 imagery can help alleviate the saturation problem and improve the accuracy of LAI estimation in a mangrove area.

Original languageEnglish (US)
Article number1060
JournalRemote Sensing
Volume9
Issue number10
DOIs
StatePublished - Oct 1 2017

Fingerprint

vegetation index
leaf area index
mangrove
artificial neural network
imagery
near infrared
machine learning
saturation
grassland
WorldView
sensor
fold
calibration

Keywords

  • Leaf area index
  • Machine learning
  • Mangrove forests
  • Red-edge band
  • Variable importance
  • Vegetation index
  • WorldView-2 imagery

ASJC Scopus subject areas

  • Earth and Planetary Sciences(all)

Cite this

Exploring the potential of world view-2 red-edge band-based vegetation indices for estimation of mangrove leaf area index with machine learning algorithms. / Zhu, Yuanhui; Liu, Kai; Liu, Lin; Myint, Soe; Wang, Shugong; Liu, Hongxing; He, Zhi.

In: Remote Sensing, Vol. 9, No. 10, 1060, 01.10.2017.

Research output: Contribution to journalArticle

@article{967a927d336646f98e1cb22460a7bb27,
title = "Exploring the potential of world view-2 red-edge band-based vegetation indices for estimation of mangrove leaf area index with machine learning algorithms",
abstract = "To accurately estimate leaf area index (LAI) in mangrove areas, the selection of appropriate models and predictor variables is critical. However, there is a major challenge in quantifying and mapping LAI using multi-spectral sensors due to the saturation effects of traditional vegetation indices (VIs) for mangrove forests. WorldView-2 (WV2) imagery has proven to be effective to estimate LAI of grasslands and forests, but the sensitivity of its vegetation indices (VIs) has been uncertain for mangrove forests. Furthermore, the single model may exhibit certain randomness and instability in model calibration and estimation accuracy. Therefore, this study aims to explore the sensitivity of WV2 VIs for estimating mangrove LAI by comparing artificial neural network regression (ANNR), support vector regression (SVR) and random forest regression (RFR). The results suggest that the RFR algorithm yields the best results (RMSE = 0.45, 14.55{\%} of the average LAI), followed by ANNR (RMSE = 0.49, 16.04{\%} of the average LAI), and then SVR (RMSE = 0.51, 16.56{\%} of the average LAI) algorithms using 5-fold cross validation (CV) using all VIs. Quantification of the variable importance shows that the VIs derived from the red-edge band consistently remain the most important contributor to LAI estimation. When the red-edge band-derived VIs are removed from the models, estimation accuracies measured in relative RMSE (RMSEr) decrease by 3.79{\%}, 2.70{\%} and 4.47{\%} for ANNR, SVR and RFR models respectively. VIs derived from red-edge band also yield better accuracy compared with other traditional bands of WV2, such as near-infrared-1 and near-infrared-2 band. Furthermore, the estimated LAI values vary significantly across different mangrove species. The study demonstrates the utility of VIs of WV2 imagery and the selected machine-learning algorithms in developing LAI models in mangrove forests. The results indicate that the red-edge band of WV2 imagery can help alleviate the saturation problem and improve the accuracy of LAI estimation in a mangrove area.",
keywords = "Leaf area index, Machine learning, Mangrove forests, Red-edge band, Variable importance, Vegetation index, WorldView-2 imagery",
author = "Yuanhui Zhu and Kai Liu and Lin Liu and Soe Myint and Shugong Wang and Hongxing Liu and Zhi He",
year = "2017",
month = "10",
day = "1",
doi = "10.3390/rs9101060",
language = "English (US)",
volume = "9",
journal = "Remote Sensing",
issn = "2072-4292",
publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",
number = "10",

}

TY - JOUR

T1 - Exploring the potential of world view-2 red-edge band-based vegetation indices for estimation of mangrove leaf area index with machine learning algorithms

AU - Zhu, Yuanhui

AU - Liu, Kai

AU - Liu, Lin

AU - Myint, Soe

AU - Wang, Shugong

AU - Liu, Hongxing

AU - He, Zhi

PY - 2017/10/1

Y1 - 2017/10/1

N2 - To accurately estimate leaf area index (LAI) in mangrove areas, the selection of appropriate models and predictor variables is critical. However, there is a major challenge in quantifying and mapping LAI using multi-spectral sensors due to the saturation effects of traditional vegetation indices (VIs) for mangrove forests. WorldView-2 (WV2) imagery has proven to be effective to estimate LAI of grasslands and forests, but the sensitivity of its vegetation indices (VIs) has been uncertain for mangrove forests. Furthermore, the single model may exhibit certain randomness and instability in model calibration and estimation accuracy. Therefore, this study aims to explore the sensitivity of WV2 VIs for estimating mangrove LAI by comparing artificial neural network regression (ANNR), support vector regression (SVR) and random forest regression (RFR). The results suggest that the RFR algorithm yields the best results (RMSE = 0.45, 14.55% of the average LAI), followed by ANNR (RMSE = 0.49, 16.04% of the average LAI), and then SVR (RMSE = 0.51, 16.56% of the average LAI) algorithms using 5-fold cross validation (CV) using all VIs. Quantification of the variable importance shows that the VIs derived from the red-edge band consistently remain the most important contributor to LAI estimation. When the red-edge band-derived VIs are removed from the models, estimation accuracies measured in relative RMSE (RMSEr) decrease by 3.79%, 2.70% and 4.47% for ANNR, SVR and RFR models respectively. VIs derived from red-edge band also yield better accuracy compared with other traditional bands of WV2, such as near-infrared-1 and near-infrared-2 band. Furthermore, the estimated LAI values vary significantly across different mangrove species. The study demonstrates the utility of VIs of WV2 imagery and the selected machine-learning algorithms in developing LAI models in mangrove forests. The results indicate that the red-edge band of WV2 imagery can help alleviate the saturation problem and improve the accuracy of LAI estimation in a mangrove area.

AB - To accurately estimate leaf area index (LAI) in mangrove areas, the selection of appropriate models and predictor variables is critical. However, there is a major challenge in quantifying and mapping LAI using multi-spectral sensors due to the saturation effects of traditional vegetation indices (VIs) for mangrove forests. WorldView-2 (WV2) imagery has proven to be effective to estimate LAI of grasslands and forests, but the sensitivity of its vegetation indices (VIs) has been uncertain for mangrove forests. Furthermore, the single model may exhibit certain randomness and instability in model calibration and estimation accuracy. Therefore, this study aims to explore the sensitivity of WV2 VIs for estimating mangrove LAI by comparing artificial neural network regression (ANNR), support vector regression (SVR) and random forest regression (RFR). The results suggest that the RFR algorithm yields the best results (RMSE = 0.45, 14.55% of the average LAI), followed by ANNR (RMSE = 0.49, 16.04% of the average LAI), and then SVR (RMSE = 0.51, 16.56% of the average LAI) algorithms using 5-fold cross validation (CV) using all VIs. Quantification of the variable importance shows that the VIs derived from the red-edge band consistently remain the most important contributor to LAI estimation. When the red-edge band-derived VIs are removed from the models, estimation accuracies measured in relative RMSE (RMSEr) decrease by 3.79%, 2.70% and 4.47% for ANNR, SVR and RFR models respectively. VIs derived from red-edge band also yield better accuracy compared with other traditional bands of WV2, such as near-infrared-1 and near-infrared-2 band. Furthermore, the estimated LAI values vary significantly across different mangrove species. The study demonstrates the utility of VIs of WV2 imagery and the selected machine-learning algorithms in developing LAI models in mangrove forests. The results indicate that the red-edge band of WV2 imagery can help alleviate the saturation problem and improve the accuracy of LAI estimation in a mangrove area.

KW - Leaf area index

KW - Machine learning

KW - Mangrove forests

KW - Red-edge band

KW - Variable importance

KW - Vegetation index

KW - WorldView-2 imagery

UR - http://www.scopus.com/inward/record.url?scp=85032880259&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85032880259&partnerID=8YFLogxK

U2 - 10.3390/rs9101060

DO - 10.3390/rs9101060

M3 - Article

VL - 9

JO - Remote Sensing

JF - Remote Sensing

SN - 2072-4292

IS - 10

M1 - 1060

ER -