Gang Yang

Missing Information Reconstruction of Remote Sensing Data: A Technical Review

Because of sensor malfunction and poor atmospheric conditions, there is usually a great deal of missing information in optical remote sensing data, which reduces the usage rate and hinders the follow-up interpretation. In the past decades, missing information reconstruction of remote sensing data has become an active research field, and a large number of algorithms have been developed. However, to the best of our knowledge, there has not, to date, been a study that has been aimed at expatiating and summarizing the current situation. This is therefore our motivation in this review. This paper provides an introduction to the principles and theories of missing information reconstruction of remote sensing data. We classify the established and emerging algorithms into four main categories, followed by a comprehensive comparison of them from both experimental and theoretical perspectives. This paper also predicts the promising future research directions.

Recovering Quantitative Remote Sensing Products Contaminated by Thick Clouds and Shadows Using Multitemporal Dictionary Learning

With regard to quantitative remote sensing products in the visible and infrared ranges, thick clouds and accompanying shadows are an inevitable source of noise. Due to the absence of adequate supporting information from the data themselves, it is a formidable challenge to accurately restore the surficial information underlying large-scale clouds. In this paper, dictionary learning is expanded into the multitemporal recovery of quantitative data contaminated by thick clouds and shadows. This paper proposes two multitemporal dictionary learning algorithms, expanding on their KSVD and Bayesian counterparts. In order to make better use of the temporal correlations, the expanded KSVD algorithm seeks an optimized temporal path, and the expanded Bayesian method adaptively weights the temporal correlations. In the experiments, the proposed algorithms are applied to a reflectance product and a land surface temperature product, and the respective advantages of the two algorithms are investigated. The results show that, from both the qualitative visual effect and the quantitative objective evaluation, the proposed methods are effective.

A Moving Weighted Harmonic Analysis Method for Reconstructing High-Quality SPOT VEGETATION NDVI Time-Series Data

Global or regional environmental change is of wide concern. Extensive studies have indicated that long-term vegetation cover change is one of the most important factors reflecting environmental change, and normalized difference vegetation index (NDVI) time-series data sets have been widely used in vegetation dynamic change monitoring. However, the significant residual effects and noise levels impede the application of NDVI time-series data in environmental change research. This study develops a novel and robust filter method, i.e., the moving weighted harmonic analysis (MWHA) method, which incorporates a moving support domain to assign the weights for all the points, making the determination of the frequency number much easier. Additionally, a four-step process flow is designed to make the data approach the upper NDVI envelope, so that the actual change in the vegetation can be detected. A total of 487 test pixels selected from SPOT VEGETATION 10-day MVC NDVI time-series data from January 1999 to December 2001 were used to illustrate the effectiveness of the new method by comparing the MWHA results with the results of another four existing methods. Finally, the long-term SPOT VEGETATION 10-day maximum-value compositing (MVC) NDVI time series for China from April 1998 to May 2014 was reconstructed by the use of the proposed method, and a test region in China was utilized to validate the effectiveness of the proposed MWHA method. All the results indicate that the reconstructed high-quality NDVI time series fits the actual growth profile of the vegetation and is suitable for use in further remote sensing applications.

A Sparse and Low-Rank Near-Isometric Linear Embedding Method for Feature Extraction in Hyperspectral Imagery Classification

A sparse and low-rank near-isometric linear embedding (SLRNILE) method has been proposed to make dimensionality reduction and extract proper features for hyperspectral imagery (HSI) classification. The SLRNILE stands on the theory of the John-Lindenstrauss lemma, and tries to estimate a sparse and low-rank projection matrix that satisfies the restricted isometric property (RIP) condition on all secants of the HSI data. The RIP condition guarantees that the desired linear mapping near-isometrically preserves nearest neighbor points of all HSI pixels. Seeking the desired mapping is then modeled into minimizing a Lagrange multipliers formulation. The alternating direction method of multipliers framework is utilized to solve the above convex program, and column generation techniques are adopted to alleviate the computation memory burden during the optimization procedure. Five experiments on three widely used HSI data sets are designed to completely test the performance of SLRNILE, and experimental results are compared against those of six state-of-the-art feature extraction methods, including principal component analysis, Laplacian eigenmaps, locality preserving projections, neighborhood preserving embedding, sparse nonnegative matrix underapproximation, and random projections. The results show that SLRNILE performs best among all the seven methods, and its computational time is longest of all but still bearable for regular users. Therefore, the SLRNILE can be a good choice for feature extraction in HSI classification.

Modelling Seasonal GWR of Daily PM2.5 with Proper Auxiliary Variables for the Yangtze River Delta

Over the past decades, regional haze episodes have frequently occurred in eastern China, especially in the Yangtze River Delta (YRD). Satellite derived Aerosol Optical Depth (AOD) has been used to retrieve the spatial coverage of PM2.5 concentrations. To improve the retrieval accuracy of the daily AOD-PM2.5 model, various auxiliary variables like meteorological or geographical factors have been adopted into the Geographically Weighted Regression (GWR) model. However, these variables are always arbitrarily selected without deep consideration of their potentially varying temporal or spatial contributions in the model performance. In this manuscript, we put forward an automatic procedure to select proper auxiliary variables from meteorological and geographical factors and obtain their optimal combinations to construct four seasonal GWR models. We employ two different schemes to comprehensively test the performance of our proposed GWR models: (1) comparison with other regular GWR models by varying the number of auxiliary variables; and (2) comparison with observed ground-level PM2.5 concentrations. The result shows that our GWR models of “AOD + 3” with three common meteorological variables generally perform better than all the other GWR models involved. Our models also show powerful prediction capabilities in PM2.5 concentrations with only slight overfitting. The determination coefficients R2 of our seasonal models are 0.8259 in spring, 0.7818 in summer, 0.8407 in autumn, and 0.7689 in winter. Also, the seasonal models in summer and autumn behave better than those in spring and winter. The comparison between seasonal and yearly models further validates the specific seasonal pattern of auxiliary variables of the GWR model in the YRD. We also stress the importance of key variables and propose a selection process in the AOD-PM2.5 model. Our work validates the significance of proper auxiliary variables in modelling the AOD-PM2.5 relationships and provides a good alternative in retrieving daily PM2.5 concentrations from remote sensing images in the YRD.

A 33-Year NPP Monitoring Study in Southwest China by the Fusion of Multi-Source Remote Sensing and Station Data

Knowledge of regional net primary productivity (NPP) is important for the systematic understanding of the global carbon cycle. In this study, multi-source data were employed to conduct a regional NPP study in southwest China, with a 33-year time span and a 1-km scale. A multi-sensor fusion framework was applied to obtain a new normalized difference vegetation index (NDVI) time series from 1982 to 2014, combining the advantages of different remote sensing datasets. As another key parameter for NPP modeling, the total solar radiation was calculated utilizing the improved Yang hybrid model (YHM), based on meteorological station data. The accuracy of the data processes is proved reliable by verification experiments. Moreover, NPP estimated by fused NDVI shows an obvious improved accuracy than that based on the original data. The spatio-temporal analysis results indicated that 67% of the study area showed an increasing NPP trend over the past three decades. The correlation between NPP and precipitation was significant heterogeneous at the monthly scale; specifically, the correlation is negative in the growing season and positive in the dry season. Meanwhile, the lagged positive correlation in the growing season and no lag in the dry season indicated the important impacts of precipitation on NPP. What is more, we found that there are three distinct stages during the variation of NPP, which were driven by different climatic factors. Significant climate warming led to a great increase of NPP from 1992 to 2002, while NPP clearly decreased during 1982–1992 and 2002–2014 due to the frequent droughts caused by the precipitation decrease.

Spatially continuous mapping of daily global ozone distribution (2004–2014) with the Aura OMI sensor

Total ozone data from the Aura Ozone Monitoring Instrument (OMI) play an important role in the monitoring of the spatial distribution and temporal change of total ozone. However, since September 2005, and especially after mid-2006, due to row anomalies in the OMI instrument, one third to one half of the OMI total ozone data has been missing. In this study, we generate a spatially continuous and daily global total ozone product (2004–2014) by quantitatively reconstructing the level-3 (gridded) total ozone data via a new two-step method, which takes full advantage of the temporal and spatial correlation characteristics. Firstly, a preliminary prediction is made based on an adaptive weighted temporal fitting method. Residual correction based on an anisotropic kriging method is then proposed to further improve the prediction accuracy. To assess the efficacy of the proposed method, a comparison of different gap filling algorithms through a series of simulated experiments was performed. On this basis, we further evaluated the proposed product with Brewer spectrophotometers’ total ozone columns. The evaluation results suggest that the proposed method outperforms the other algorithms, and its product is better able to capture total ozone variation than the MERRA-2 assimilated ozone product. The total ozone product produced in this study can be freely downloaded from http://sendimage.whu.edu.cn/send-resource-download/.

A Probabilistic Weighted Archetypal Analysis Method with Earth Mover’s Distance for Endmember Extraction from Hyperspectral Imagery

A Probabilistic Weighted Archetypal Analysis method with Earth Mover’s Distance (PWAA-EMD) is proposed to extract endmembers from hyperspectral imagery (HSI). The PWAA-EMD first utilizes the EMD dissimilarity matrix to weight the coefficient matrix in the regular Archetypal Analysis (AA). The EMD metric considers manifold structures of spectral signatures in the HSI data and could better quantify the dissimilarity features among pairwise pixels. Second, the PWAA-EMD adopts the Bayesian framework and formulates the improved AA into a probabilistic inference problem by maximizing a joint posterior density. Third, the optimization problem is solved by the iterative multiplicative update scheme, with a careful initialization from the two-stage algorithm and the proper endmembers are finally obtained. The synthetic and real Cuprite Hyperspectral datasets are utilized to verify the performance of PWAA-EMD and five popular methods are implemented to make comparisons. The results show that PWAA-EMD surpasses all the five methods in the average results of spectral angle distance (SAD) and root-mean-square-error (RMSE). Especially, the PWAA-EMD obtains more accurate estimation than AA in almost all the classes of endmembers including two similar ones. Therefore, the PWAA-EMD could be an alternative choice for endmember extraction on the hyperspectral data.

Hydrological Analysis Using Satellite Remote Sensing Big Data and CREST Model

Hydrological modeling significantly contributes to the understanding of catchment water balance and water resource management and mitigates negative impacts of flooding. Considering the advantages of satellite remote sensing big data and the coupled routing and excess storage (CREST) model, this paper investigates the hydrological modeling in the Shehong basin during 2006–2013. The results show that humid Shehong basin has main rainfalls in summer (From May to September). For the monthly average rainfall and streamflow, there is a remarkable increase (+52%) in discharge and a smaller increase (+18%) in rainfall in the second period (2010–2013) relative to the first period (2006–2009). The CREST model was calibrated using China gauge-based daily precipitation analysis for the period of 2006–2009, followed by a favorable performance with Nash-Sutcliffe coefficient efficiency (NSCE) of 0.77, correlation coefficient (CC) up to 0.88 and −11% Bias. The model validation shows an error metric with NSCE of 0.74, CC of 0.87 and −11.7% Bias. In terms of water balance modeling results at Shehong basin, the runoff and rainfall estimates from CREST model coincide well with the gauge observations, indicating the model captures the appropriate signature of soil moisture variability. Therefore, the satellite-based precipitation product is feasible in hydrological prediction, and the CREST models the interaction between surface and subsurface water flow process in the Shehong basin.