Home > Research > Publications & Outputs > Ensembles of multiple spectral water indices fo...

Electronic data

  • Manuscript_IJAEOG_Accepted

    Accepted author manuscript, 4.77 MB, PDF document

    Available under license: CC BY: Creative Commons Attribution 4.0 International License

Links

Text available via DOI:

View graph of relations

Ensembles of multiple spectral water indices for improving surface water classification

Research output: Contribution to journalJournal articlepeer-review

Published

Standard

Ensembles of multiple spectral water indices for improving surface water classification. / Wen, Zhaofei; Zhang, Ce; Shao, Guofan; Wu, Shengjun; Atkinson, Peter.

In: International Journal of Applied Earth Observation and Geoinformation, Vol. 96, 102278, 01.04.2021.

Research output: Contribution to journalJournal articlepeer-review

Harvard

Wen, Z, Zhang, C, Shao, G, Wu, S & Atkinson, P 2021, 'Ensembles of multiple spectral water indices for improving surface water classification', International Journal of Applied Earth Observation and Geoinformation, vol. 96, 102278. https://doi.org/10.1016/j.jag.2020.102278

APA

Wen, Z., Zhang, C., Shao, G., Wu, S., & Atkinson, P. (2021). Ensembles of multiple spectral water indices for improving surface water classification. International Journal of Applied Earth Observation and Geoinformation, 96, [102278]. https://doi.org/10.1016/j.jag.2020.102278

Vancouver

Wen Z, Zhang C, Shao G, Wu S, Atkinson P. Ensembles of multiple spectral water indices for improving surface water classification. International Journal of Applied Earth Observation and Geoinformation. 2021 Apr 1;96. 102278. https://doi.org/10.1016/j.jag.2020.102278

Author

Wen, Zhaofei ; Zhang, Ce ; Shao, Guofan ; Wu, Shengjun ; Atkinson, Peter. / Ensembles of multiple spectral water indices for improving surface water classification. In: International Journal of Applied Earth Observation and Geoinformation. 2021 ; Vol. 96.

Bibtex

@article{c48c932766b045ef9a7a774cfba2939b,
title = "Ensembles of multiple spectral water indices for improving surface water classification",
abstract = "Mapping surface water distribution and its dynamics over various environments with robust methods is essential for managing water resources and supporting water-related policy design. Thresholding Single Water Index image (TSWI) with a fixed threshold is a common way of using water index (WI) for mapping water for it is easy to use and could obtain acceptable accuracies in many applications. As more and more WIs are available and each has its distinct merits, the real-world application of TSWI, however, often face two practical concerns: (1) selection of an appropriate WI, and (2) determination of an optimal threshold for a given WI. These two issues are problematic for many users who rely either on trial-and-error procedures that are time-consuming or on their personal preferences that are somewhat subjective. To better deal with these two practical concerns, an alternative way of using WIs is suggested here by transforming the current paradigm into a simple but robust ensemble approach called Collaborative Decision-making with Water Indices (CDWI). A total of 145 subsite images (900  900 m) from 22 Landsat-8 OLI scenes that covering various water-land environments around the world were used to assess the performance of TSWI and the CDWI. Five benchmark WIs were adopted in five TSWI methods and CDWI method: Normalized Difference Water Index (NDWI), the Modified NDWI (MNDWI), the Automated Water Extraction Indices without considering (AWEI0) and with considering (AWEI1) shadows, and the state-of-the-art 2015 water index (WI2015). Two aspects of performance were analyzed: comparing their accuracies (indicated by both F1-scores and Youden{\textquoteright}s Index) over various environments and comparing their accuracy sensitivities to threshold. The results demonstrate that CDWI produced higher accuracies than the other five TSWI methods for most application cases. Particularly, more samples (indicated by percentage) produced higher F1-scores by CDWI than the other five TSWI methods, i.e. 67% (CDWI) vs. 15% (TSWINDWI), 54% (CDWI) vs. 22% (TSWIMNDWI), 42% (CDWI) vs. 12% (TSWIAWEI0), 57% (CDWI) vs. 17% (TSWIAWEI1), and 34% (CDWI) vs. 12% (TSWIWI2015). Moreover, the F1-score of the CDWI is much less sensitive to the change of thresholds compared with that of the other five TSWI methods. These important benefits of CDWI make it a robust approach for mapping water. The uncertainty of CDWI method was thoroughly discussed and a general guidance (or look-up-table) for selecting WIs was also suggested. The underlying framework of CDWI could be readily generalizable and applicable to other satellite sensor images, such as Landsat TM/ETM+, MODIS, and Sentinel-2 images.",
keywords = "Water index, Threshold, Integrated decision making, Mixed pixels, MNDWI",
author = "Zhaofei Wen and Ce Zhang and Guofan Shao and Shengjun Wu and Peter Atkinson",
year = "2021",
month = apr,
day = "1",
doi = "10.1016/j.jag.2020.102278",
language = "English",
volume = "96",
journal = "International Journal of Applied Earth Observation and Geoinformation",
issn = "0303-2434",
publisher = "International Institute for Aerial Survey and Earth Sciences",

}

RIS

TY - JOUR

T1 - Ensembles of multiple spectral water indices for improving surface water classification

AU - Wen, Zhaofei

AU - Zhang, Ce

AU - Shao, Guofan

AU - Wu, Shengjun

AU - Atkinson, Peter

PY - 2021/4/1

Y1 - 2021/4/1

N2 - Mapping surface water distribution and its dynamics over various environments with robust methods is essential for managing water resources and supporting water-related policy design. Thresholding Single Water Index image (TSWI) with a fixed threshold is a common way of using water index (WI) for mapping water for it is easy to use and could obtain acceptable accuracies in many applications. As more and more WIs are available and each has its distinct merits, the real-world application of TSWI, however, often face two practical concerns: (1) selection of an appropriate WI, and (2) determination of an optimal threshold for a given WI. These two issues are problematic for many users who rely either on trial-and-error procedures that are time-consuming or on their personal preferences that are somewhat subjective. To better deal with these two practical concerns, an alternative way of using WIs is suggested here by transforming the current paradigm into a simple but robust ensemble approach called Collaborative Decision-making with Water Indices (CDWI). A total of 145 subsite images (900  900 m) from 22 Landsat-8 OLI scenes that covering various water-land environments around the world were used to assess the performance of TSWI and the CDWI. Five benchmark WIs were adopted in five TSWI methods and CDWI method: Normalized Difference Water Index (NDWI), the Modified NDWI (MNDWI), the Automated Water Extraction Indices without considering (AWEI0) and with considering (AWEI1) shadows, and the state-of-the-art 2015 water index (WI2015). Two aspects of performance were analyzed: comparing their accuracies (indicated by both F1-scores and Youden’s Index) over various environments and comparing their accuracy sensitivities to threshold. The results demonstrate that CDWI produced higher accuracies than the other five TSWI methods for most application cases. Particularly, more samples (indicated by percentage) produced higher F1-scores by CDWI than the other five TSWI methods, i.e. 67% (CDWI) vs. 15% (TSWINDWI), 54% (CDWI) vs. 22% (TSWIMNDWI), 42% (CDWI) vs. 12% (TSWIAWEI0), 57% (CDWI) vs. 17% (TSWIAWEI1), and 34% (CDWI) vs. 12% (TSWIWI2015). Moreover, the F1-score of the CDWI is much less sensitive to the change of thresholds compared with that of the other five TSWI methods. These important benefits of CDWI make it a robust approach for mapping water. The uncertainty of CDWI method was thoroughly discussed and a general guidance (or look-up-table) for selecting WIs was also suggested. The underlying framework of CDWI could be readily generalizable and applicable to other satellite sensor images, such as Landsat TM/ETM+, MODIS, and Sentinel-2 images.

AB - Mapping surface water distribution and its dynamics over various environments with robust methods is essential for managing water resources and supporting water-related policy design. Thresholding Single Water Index image (TSWI) with a fixed threshold is a common way of using water index (WI) for mapping water for it is easy to use and could obtain acceptable accuracies in many applications. As more and more WIs are available and each has its distinct merits, the real-world application of TSWI, however, often face two practical concerns: (1) selection of an appropriate WI, and (2) determination of an optimal threshold for a given WI. These two issues are problematic for many users who rely either on trial-and-error procedures that are time-consuming or on their personal preferences that are somewhat subjective. To better deal with these two practical concerns, an alternative way of using WIs is suggested here by transforming the current paradigm into a simple but robust ensemble approach called Collaborative Decision-making with Water Indices (CDWI). A total of 145 subsite images (900  900 m) from 22 Landsat-8 OLI scenes that covering various water-land environments around the world were used to assess the performance of TSWI and the CDWI. Five benchmark WIs were adopted in five TSWI methods and CDWI method: Normalized Difference Water Index (NDWI), the Modified NDWI (MNDWI), the Automated Water Extraction Indices without considering (AWEI0) and with considering (AWEI1) shadows, and the state-of-the-art 2015 water index (WI2015). Two aspects of performance were analyzed: comparing their accuracies (indicated by both F1-scores and Youden’s Index) over various environments and comparing their accuracy sensitivities to threshold. The results demonstrate that CDWI produced higher accuracies than the other five TSWI methods for most application cases. Particularly, more samples (indicated by percentage) produced higher F1-scores by CDWI than the other five TSWI methods, i.e. 67% (CDWI) vs. 15% (TSWINDWI), 54% (CDWI) vs. 22% (TSWIMNDWI), 42% (CDWI) vs. 12% (TSWIAWEI0), 57% (CDWI) vs. 17% (TSWIAWEI1), and 34% (CDWI) vs. 12% (TSWIWI2015). Moreover, the F1-score of the CDWI is much less sensitive to the change of thresholds compared with that of the other five TSWI methods. These important benefits of CDWI make it a robust approach for mapping water. The uncertainty of CDWI method was thoroughly discussed and a general guidance (or look-up-table) for selecting WIs was also suggested. The underlying framework of CDWI could be readily generalizable and applicable to other satellite sensor images, such as Landsat TM/ETM+, MODIS, and Sentinel-2 images.

KW - Water index

KW - Threshold

KW - Integrated decision making

KW - Mixed pixels

KW - MNDWI

U2 - 10.1016/j.jag.2020.102278

DO - 10.1016/j.jag.2020.102278

M3 - Journal article

VL - 96

JO - International Journal of Applied Earth Observation and Geoinformation

JF - International Journal of Applied Earth Observation and Geoinformation

SN - 0303-2434

M1 - 102278

ER -