Small baseline InSAR time series analysis: Unwrapping error correction and noise reduction

Zhang Yunjun, Heresh Fattahi, Falk Amelung

Research output: Contribution to journalReview article

Abstract

We present a review of small baseline interferometric synthetic aperture radar (InSAR) time series analysis with a new processing workflow and software implemented in Python, named MintPy (https://github.com/insarlab/MintPy). The time series analysis is formulated as a weighted least squares inversion. The inversion is unbiased for a fully connected network of interferograms without multiple subsets, such as provided by modern SAR satellites with small orbital tube and short revisit time. In the routine workflow, we first invert the interferogram stack for the raw phase time-series, then correct for the deterministic phase components: the tropospheric delay (using global atmospheric models or the delay-elevation ratio), the topographic residual and/or phase ramp, to obtain the noise-reduced displacement time-series. Next, we estimate the average velocity excluding noisy SAR acquisitions, which are identified using an outlier detection method based on the root mean square of the residual phase. The routine workflow includes three new methods to correct or exclude phase-unwrapping errors for two-dimensional algorithms: (i) the bridging method connecting reliable regions with minimum spanning tree bridges (particularly suitable for islands), (ii) the phase closure method exploiting the conservativeness of the integer ambiguity of interferogram triplets (well suited for highly redundant networks), and (iii) coherence-based network modification to identify and exclude interferograms with remaining coherent phase-unwrapping errors. We apply the routine workflow to the Galápagos volcanoes using Sentinel-1 and ALOS-1 data, assess the qualities of the essential steps in the workflow and compare the results with independent GPS measurements. We discuss the advantages and limitations of temporal coherence as a reliability measure, evaluate the impact of network redundancy on the precision and reliability of the InSAR measurements and its practical implication for interferometric pairs selection. A comparison with another open-source time series analysis software demonstrates the superior performance of the approach implemented in MintPy in challenging scenarios.

Original languageEnglish (US)
Article number104331
JournalComputers and Geosciences
Volume133
DOIs
StatePublished - Dec 2019
Externally publishedYes

Fingerprint

error correction
Time series analysis
time series analysis
Error correction
Synthetic aperture radar
Noise abatement
synthetic aperture radar
Time series
Radar measurement
Volcanoes
time series
software
ALOS
Redundancy
Global positioning system
detection method
outlier
Satellites
GPS
volcano

Keywords

  • Galápagos
  • InSAR
  • Phase correction
  • Phase-unwrapping error
  • Time series analysis

ASJC Scopus subject areas

  • Information Systems
  • Computers in Earth Sciences

Cite this

Small baseline InSAR time series analysis : Unwrapping error correction and noise reduction. / Yunjun, Zhang; Fattahi, Heresh; Amelung, Falk.

In: Computers and Geosciences, Vol. 133, 104331, 12.2019.

Research output: Contribution to journalReview article

@article{99a78628ac09457e9db12131389d5acf,
title = "Small baseline InSAR time series analysis: Unwrapping error correction and noise reduction",
abstract = "We present a review of small baseline interferometric synthetic aperture radar (InSAR) time series analysis with a new processing workflow and software implemented in Python, named MintPy (https://github.com/insarlab/MintPy). The time series analysis is formulated as a weighted least squares inversion. The inversion is unbiased for a fully connected network of interferograms without multiple subsets, such as provided by modern SAR satellites with small orbital tube and short revisit time. In the routine workflow, we first invert the interferogram stack for the raw phase time-series, then correct for the deterministic phase components: the tropospheric delay (using global atmospheric models or the delay-elevation ratio), the topographic residual and/or phase ramp, to obtain the noise-reduced displacement time-series. Next, we estimate the average velocity excluding noisy SAR acquisitions, which are identified using an outlier detection method based on the root mean square of the residual phase. The routine workflow includes three new methods to correct or exclude phase-unwrapping errors for two-dimensional algorithms: (i) the bridging method connecting reliable regions with minimum spanning tree bridges (particularly suitable for islands), (ii) the phase closure method exploiting the conservativeness of the integer ambiguity of interferogram triplets (well suited for highly redundant networks), and (iii) coherence-based network modification to identify and exclude interferograms with remaining coherent phase-unwrapping errors. We apply the routine workflow to the Gal{\'a}pagos volcanoes using Sentinel-1 and ALOS-1 data, assess the qualities of the essential steps in the workflow and compare the results with independent GPS measurements. We discuss the advantages and limitations of temporal coherence as a reliability measure, evaluate the impact of network redundancy on the precision and reliability of the InSAR measurements and its practical implication for interferometric pairs selection. A comparison with another open-source time series analysis software demonstrates the superior performance of the approach implemented in MintPy in challenging scenarios.",
keywords = "Gal{\'a}pagos, InSAR, Phase correction, Phase-unwrapping error, Time series analysis",
author = "Zhang Yunjun and Heresh Fattahi and Falk Amelung",
year = "2019",
month = "12",
doi = "10.1016/j.cageo.2019.104331",
language = "English (US)",
volume = "133",
journal = "Computers and Geosciences",
issn = "0098-3004",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - Small baseline InSAR time series analysis

T2 - Unwrapping error correction and noise reduction

AU - Yunjun, Zhang

AU - Fattahi, Heresh

AU - Amelung, Falk

PY - 2019/12

Y1 - 2019/12

N2 - We present a review of small baseline interferometric synthetic aperture radar (InSAR) time series analysis with a new processing workflow and software implemented in Python, named MintPy (https://github.com/insarlab/MintPy). The time series analysis is formulated as a weighted least squares inversion. The inversion is unbiased for a fully connected network of interferograms without multiple subsets, such as provided by modern SAR satellites with small orbital tube and short revisit time. In the routine workflow, we first invert the interferogram stack for the raw phase time-series, then correct for the deterministic phase components: the tropospheric delay (using global atmospheric models or the delay-elevation ratio), the topographic residual and/or phase ramp, to obtain the noise-reduced displacement time-series. Next, we estimate the average velocity excluding noisy SAR acquisitions, which are identified using an outlier detection method based on the root mean square of the residual phase. The routine workflow includes three new methods to correct or exclude phase-unwrapping errors for two-dimensional algorithms: (i) the bridging method connecting reliable regions with minimum spanning tree bridges (particularly suitable for islands), (ii) the phase closure method exploiting the conservativeness of the integer ambiguity of interferogram triplets (well suited for highly redundant networks), and (iii) coherence-based network modification to identify and exclude interferograms with remaining coherent phase-unwrapping errors. We apply the routine workflow to the Galápagos volcanoes using Sentinel-1 and ALOS-1 data, assess the qualities of the essential steps in the workflow and compare the results with independent GPS measurements. We discuss the advantages and limitations of temporal coherence as a reliability measure, evaluate the impact of network redundancy on the precision and reliability of the InSAR measurements and its practical implication for interferometric pairs selection. A comparison with another open-source time series analysis software demonstrates the superior performance of the approach implemented in MintPy in challenging scenarios.

AB - We present a review of small baseline interferometric synthetic aperture radar (InSAR) time series analysis with a new processing workflow and software implemented in Python, named MintPy (https://github.com/insarlab/MintPy). The time series analysis is formulated as a weighted least squares inversion. The inversion is unbiased for a fully connected network of interferograms without multiple subsets, such as provided by modern SAR satellites with small orbital tube and short revisit time. In the routine workflow, we first invert the interferogram stack for the raw phase time-series, then correct for the deterministic phase components: the tropospheric delay (using global atmospheric models or the delay-elevation ratio), the topographic residual and/or phase ramp, to obtain the noise-reduced displacement time-series. Next, we estimate the average velocity excluding noisy SAR acquisitions, which are identified using an outlier detection method based on the root mean square of the residual phase. The routine workflow includes three new methods to correct or exclude phase-unwrapping errors for two-dimensional algorithms: (i) the bridging method connecting reliable regions with minimum spanning tree bridges (particularly suitable for islands), (ii) the phase closure method exploiting the conservativeness of the integer ambiguity of interferogram triplets (well suited for highly redundant networks), and (iii) coherence-based network modification to identify and exclude interferograms with remaining coherent phase-unwrapping errors. We apply the routine workflow to the Galápagos volcanoes using Sentinel-1 and ALOS-1 data, assess the qualities of the essential steps in the workflow and compare the results with independent GPS measurements. We discuss the advantages and limitations of temporal coherence as a reliability measure, evaluate the impact of network redundancy on the precision and reliability of the InSAR measurements and its practical implication for interferometric pairs selection. A comparison with another open-source time series analysis software demonstrates the superior performance of the approach implemented in MintPy in challenging scenarios.

KW - Galápagos

KW - InSAR

KW - Phase correction

KW - Phase-unwrapping error

KW - Time series analysis

UR - http://www.scopus.com/inward/record.url?scp=85072919630&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85072919630&partnerID=8YFLogxK

U2 - 10.1016/j.cageo.2019.104331

DO - 10.1016/j.cageo.2019.104331

M3 - Review article

AN - SCOPUS:85072919630

VL - 133

JO - Computers and Geosciences

JF - Computers and Geosciences

SN - 0098-3004

M1 - 104331

ER -