4.7 Article

Soil Moisture Estimation Using Sentinel-1/-2 Imagery Coupled With CycleGAN for Time-Series Gap Filing

Journal

Publisher

IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TGRS.2021.3134127

Keywords

Soil moisture; Feature extraction; Satellites; Data models; Predictive models; Training; Soil measurements; Agriculture; generative adversarial networks (GANs); machine learning (ML); Sentinel-1; Sentinel-2; soil moisture (SM); unsupervised domain adaptation

Funding

  1. Space Research and Innovation Network for Technology (SPRINT) [1243832]
  2. Research Fund of the Istanbul Technical University [MGA-2021-43018]

Ask authors/readers for more resources

This study explores the possibility of using freely available Sentinel-1 and Sentinel-2 earth observation data for the simultaneous prediction of soil moisture content (SMC) using a cycle-consistent adversarial network (CycleGAN) for time-series gap filling. The proposed methodology learns the latent low-dimensional representation of satellite images and then builds a machine learning model on top of these representations to predict SMC. Experimental results show that the proposed method outperforms existing state-of-the-art methods for filling gaps in optical and synthetic-aperture radar (SAR) images.
Fast soil moisture content (SMC) mapping is necessary to support water resource management and to understand crop growth, quality, and yield. Therefore, earth observation (EO) plays a key role due to its ability of almost real-time monitoring of large areas at a low cost. This study aimed to explore the possibility of taking advantage of freely available Sentinel-1 (S1) and Sentinel-2 (S2) EO data for the simultaneous prediction of SMC with cycle-consistent adversarial network (CycleGAN) for time-series gap filling. The proposed methodology, first, learns latent low-dimensional representation of the satellite images, then learns a simple machine learning (ML) model on top of these representations. To evaluate the methodology, a series of vineyards, located in South Australia & x2019;s Eden valley are chosen. Specifically, we presented an efficient framework for extracting latent features from S1 and S2 imagery. We showed how one could use S1 to S2 feature translation based on CycleGAN using S1 and S2 time series when there are missing images acquired over an area of interest. The resulting data in our study is then used to fill gaps in time-series data. We used the resulting latent representations to predict SMC with various ML tools. In the experiments, CycleGAN and the autoencoders were trained with data randomly chosen around the site of interest, so we could augment the existing dataset. The best performance was demonstrated with random forest (RF) algorithm, whereas linear regression model demonstrated significant overfitting. The experiments demonstrate that the proposed methodology outperforms the compared state-of-the-art methods if there are missing optical and synthetic-aperture radar (SAR) images.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.7
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available