Potential of Recurrence Metrics from Sentinel-1 Time Series for Deforestation Mapping
The REDD+ framework requires accurate estimates of deforestation. These are derived by ground measurements supported by methods based on remote sensing data to automatically detect and delineate deforestations over large areas. In particular, in the tropics, optical data is seldom available due to c...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2020-01-01
|
Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/9186165/ |
id |
doaj-cd844fa489544c589a392df79761df0a |
---|---|
record_format |
Article |
spelling |
doaj-cd844fa489544c589a392df79761df0a2021-06-03T23:03:10ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352020-01-01135233524010.1109/JSTARS.2020.30193339186165Potential of Recurrence Metrics from Sentinel-1 Time Series for Deforestation MappingFelix Cremer0https://orcid.org/0000-0001-8659-4361Mikhail Urbazaev1https://orcid.org/0000-0002-0327-6278Jose Cortes2https://orcid.org/0000-0003-2567-8689John Truckenbrodt3https://orcid.org/0000-0002-7259-101XChristiane Schmullius4Christian Thiel5https://orcid.org/0000-0001-5144-8145German Aerospace Center, Institute of Data Science, Jena, GermanyChair of Earth Observation, Friedrich-Schiller University, Jena, GermanyGIScience Group, Friedrich-Schiller University, Jena, GermanyGerman Aerospace Center, Institute of Data Science, Jena, GermanyChair of Earth Observation, Friedrich-Schiller University, Jena, GermanyGerman Aerospace Center, Institute of Data Science, Jena, GermanyThe REDD+ framework requires accurate estimates of deforestation. These are derived by ground measurements supported by methods based on remote sensing data to automatically detect and delineate deforestations over large areas. In particular, in the tropics, optical data is seldom available due to cloud cover. As synthetic aperture radar (SAR) data overcomes this limitation, we performed a separability analysis of two statistical metrics based on the Sentinel-1 SAR backscatter over forested and deforested areas. We compared the range between the 5th and 95th temporal percentiles (PRange) and the recurrence quantification analysis (RQA) Trend metric. Unlike the PRange, the RQA Trend considers the temporal order of the SAR data acquisitions, thus contrasting between dropping backscatter signals and yearly seasonalities. This enables the estimation of the timing of deforestation events. We assessed the impact of polarization, acquisition orbit, and despeckling on the separability between forested and deforested areas and between different deforestation timings for two test sites in Mexico. We found that the choice of the orbit impacts the detectability of deforestation. In all cases, VH data showed a higher separability between forest and deforestations than VV data. The PRange slightly outperformed RQA Trend in the separation between forest and deforestation. However, the RQA Trend exceeded the PRange in the separation between different deforestation timings. In this study, C-Band backscatter data was used, although it is commonly not considered as the most suitable SAR dataset for forestry applications. Nevertheless, our approach shows the potential of dense C-Band backscatter time series to support the REDD+ framework.https://ieeexplore.ieee.org/document/9186165/Forestryradar remote sensingsynthetic aperture radartime series analysis |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Felix Cremer Mikhail Urbazaev Jose Cortes John Truckenbrodt Christiane Schmullius Christian Thiel |
spellingShingle |
Felix Cremer Mikhail Urbazaev Jose Cortes John Truckenbrodt Christiane Schmullius Christian Thiel Potential of Recurrence Metrics from Sentinel-1 Time Series for Deforestation Mapping IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Forestry radar remote sensing synthetic aperture radar time series analysis |
author_facet |
Felix Cremer Mikhail Urbazaev Jose Cortes John Truckenbrodt Christiane Schmullius Christian Thiel |
author_sort |
Felix Cremer |
title |
Potential of Recurrence Metrics from Sentinel-1 Time Series for Deforestation Mapping |
title_short |
Potential of Recurrence Metrics from Sentinel-1 Time Series for Deforestation Mapping |
title_full |
Potential of Recurrence Metrics from Sentinel-1 Time Series for Deforestation Mapping |
title_fullStr |
Potential of Recurrence Metrics from Sentinel-1 Time Series for Deforestation Mapping |
title_full_unstemmed |
Potential of Recurrence Metrics from Sentinel-1 Time Series for Deforestation Mapping |
title_sort |
potential of recurrence metrics from sentinel-1 time series for deforestation mapping |
publisher |
IEEE |
series |
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
issn |
2151-1535 |
publishDate |
2020-01-01 |
description |
The REDD+ framework requires accurate estimates of deforestation. These are derived by ground measurements supported by methods based on remote sensing data to automatically detect and delineate deforestations over large areas. In particular, in the tropics, optical data is seldom available due to cloud cover. As synthetic aperture radar (SAR) data overcomes this limitation, we performed a separability analysis of two statistical metrics based on the Sentinel-1 SAR backscatter over forested and deforested areas. We compared the range between the 5th and 95th temporal percentiles (PRange) and the recurrence quantification analysis (RQA) Trend metric. Unlike the PRange, the RQA Trend considers the temporal order of the SAR data acquisitions, thus contrasting between dropping backscatter signals and yearly seasonalities. This enables the estimation of the timing of deforestation events. We assessed the impact of polarization, acquisition orbit, and despeckling on the separability between forested and deforested areas and between different deforestation timings for two test sites in Mexico. We found that the choice of the orbit impacts the detectability of deforestation. In all cases, VH data showed a higher separability between forest and deforestations than VV data. The PRange slightly outperformed RQA Trend in the separation between forest and deforestation. However, the RQA Trend exceeded the PRange in the separation between different deforestation timings. In this study, C-Band backscatter data was used, although it is commonly not considered as the most suitable SAR dataset for forestry applications. Nevertheless, our approach shows the potential of dense C-Band backscatter time series to support the REDD+ framework. |
topic |
Forestry radar remote sensing synthetic aperture radar time series analysis |
url |
https://ieeexplore.ieee.org/document/9186165/ |
work_keys_str_mv |
AT felixcremer potentialofrecurrencemetricsfromsentinel1timeseriesfordeforestationmapping AT mikhailurbazaev potentialofrecurrencemetricsfromsentinel1timeseriesfordeforestationmapping AT josecortes potentialofrecurrencemetricsfromsentinel1timeseriesfordeforestationmapping AT johntruckenbrodt potentialofrecurrencemetricsfromsentinel1timeseriesfordeforestationmapping AT christianeschmullius potentialofrecurrencemetricsfromsentinel1timeseriesfordeforestationmapping AT christianthiel potentialofrecurrencemetricsfromsentinel1timeseriesfordeforestationmapping |
_version_ |
1721398830335787008 |