A Machine Learning Approach for Remote Sensing Data Gap-Filling with Open-Source Implementation: An Example Regarding Land Surface Temperature, Surface Albedo and NDVI

Reconstruction of Gap-Free Land Surface Temperature at a 100 m Spatial Resolution from Multidimensional Data: A Case in Wuhan, China

Land surface temperatures (LST) are vital parameters in land surface-atmosphere interactions. Constrained by technology and atmospheric interferences, LST retrievals from various satellite sensors usually return missing data, thus negatively impacting analyses. Reconstructing missing data is important for acquiring gap-free datasets. However, the current reconstruction methods are limited for maintaining spatial details and high accuracies. We developed a new gap-free algorithm termed the spatial feature-considered random forest regression (SFRFR) model; it builds stable nonlinear relationships to connect the LST with related parameters, including terrain elements, land coverage types, spectral indexes, surface reflectance data, and the spatial feature of the LST, to reconstruct the missing LST data. The SFRFR model reconstructed gap-free LST data retrieved from the Landsat 8 satellite on 27 July 2017 in Wuhan. The results show that the SFRFR model exhibits the best performance according to the various evaluation metrics among the SFRFR, random forest regression and spline interpolation, with a coefficient of determination (R²) reaching 0.96, root-mean-square error (RMSE) of 0.55, and mean absolute error (MAE) of 0.55. Then, we reconstructed gap-free LST data gathered in Wuhan from 2016 to 2021 to analyze urban thermal environment changes and found that 2020 presented the coolest temperatures. The SFRFR model still displayed satisfactory results, with an average R² of 0.91 and an MAE of 0.63. We further discuss and discover the factors affecting the visual performance of SFRFR and identify the research priority to circumvent these disadvantages. Overall, this study provides a simple, practical method for acquiring gap-free LST data to help us better understand the spatiotemporal LST variation process.

Study on Spatial and Temporal Characteristics of Surface Albedo at the Northern Edge of the Badain Jaran Desert Based on C + STNLFFM Model

Obtaining surface albedo data with high spatial and temporal resolution is essential for measuring the factors, effects, and change mechanisms of regional land-atmosphere interactions in deserts. In order to obtain surface albedo data with higher accuracy and better applicability in deserts, we used MODIS and OLI as data sources, and calculated the daily surface albedo data, with a spatial resolution of 30 m, of Guaizi Lake at the northern edge of the Badain Jaran Desert in 2016, using the Spatial and Temporal Non-Local Filter-based Fusion Model (STNLFFM) and topographical correction model (C model). We then compared the results of STNLFFM and C + STNLFFM for fusion accuracy, and for spatial and temporal distribution differences in surface albedo over different underlying surfaces. The results indicated that, compared with STNLFFM surface albedo and MODIS surface albedo, the relative error of C + STNLFFM surface albedo decreased by 2.34% and 3.57%, respectively. C + STNLFFM can improve poor applicability of MODIS in winter, and better responds to the changes in the measured value over a short time range. After the correction of the C model, the spatial difference in surface albedo over different underlying surfaces was enhanced, and the spatial differences in surface albedo between shifting dunes and semi-shifting dunes, fixed dunes and saline-alkali land, and the Gobi and saline-alkali land were significant. C + STNLFFM maintained the spatial and temporal distribution characteristics of STNLFFM surface albedo, but the increase in regional aerosol concentration and thickness caused by frequent dust storms weakened the spatial difference in surface albedo over different underlying surfaces in March, which led to the overcorrection of the C model.

Surface urban heat island detected by all-weather satellite land surface temperature

Since the existing satellite thermal infrared (TIR) land surface temperature (LST) is susceptible to cloud contamination and other factors, surface urban heat island (SUHI) studies based on TIR LST are limited to clear-sky conditions and are not representative of SUHI under all-weather conditions, which result in a possible clear-sky bias for SUHI. This study introduces a newly released 1-km all-weather LST product (TRIMS LST), which is spatiotemporally seamless, to investigate the real SUHI under all-weather conditions for five megacities (i.e. Harbin, Beijing, Shanghai, Guangzhou, and Chengdu) in China. Firstly, this study compares TRIMS SUHI with MODIS SUHI under clear-sky, partial-cloudy, and cloudy conditions. Secondly, the extent of the influence of cloudiness on SUHI is quantified. Finally, the monthly TRIMS SUHI is used to analyze the clear-sky bias that is caused by using only clear-sky data for the SUHI. Results indicate that (i) the absence of pixel data leads to negative offsets in the SUHI intensities calculated by MODIS LST, and these offsets expand gradually with increases in the number of missing-pixel data, causing the maximum offset to reach -1.83 °C under cloudy conditions in Chengdu; (ii) cloud can mitigate the SUHI for most cities: when the cloud coverage in Guangzhou reaches 90-100%, the daytime SUHI intensity decreases from 2.66 °C for clear-sky conditions to 1.70 °C; the mitigating effect differs at daytime and nighttime; and (iii) clear-sky bias varies significantly across climate zones and seasons, with a varying range of -1.6-1.2 °C for the five selected cities.

Spectral Temporal Information for Missing Data Reconstruction (STIMDR) of Landsat Reflectance Time Series

The number of Landsat time-series applications has grown substantially because of its approximately 50-year history and relatively high spatial resolution for observing long term changes in the Earth’s surface. However, missing observations (i.e., gaps) caused by clouds and cloud shadows, orbit and sensing geometry, and sensor issues have broadly limited the development of Landsat time-series applications. Due to the large area and temporal and spatial irregularity of time-series gaps, it is difficult to find an efficient and highly precise method to fill them. The Missing Observation Prediction based on Spectral-Temporal Metrics (MOPSTM) method has been proposed and delivered good performance in filling large-area gaps of single-date Landsat images. However, it can be less practical for a time series longer than one year due to the lack of mechanics that exclude dissimilar data in time series (e.g., different phenology or changes in land cover). To solve this problem, this study proposes a new gap-filling method, Spectral Temporal Information for Missing Data Reconstruction (STIMDR), and examines its performance in Landsat reflectance time series. Two groups of experiments, including 2000 × 2000 pixel Landsat single-date images and Landsat time series acquired from four sites (Kenya, Finland, Germany, and China), were performed to test the new method. We simulated artificial gaps to evaluate predicted pixel values with real observations. Quantitative and qualitative evaluations of gap-filled images through comparisons with other state-of-the-art methods confirmed the more robust and accurate performance of the proposed method. In addition, the proposed method was also able to fill gaps contaminated by extreme cloud cover for a period (e.g., winter in high-latitude areas). A down-stream task of random forest supervised classification through both gap-filled simulated datasets and the original valid datasets verified that STIMDR-generated products are relevant to the user community for land cover applications.

Gap-Filling Eddy Covariance Latent Heat Flux: Inter-Comparison of Four Machine Learning Model Predictions and Uncertainties in Forest Ecosystem

Environmental monitoring using satellite remote sensing is challenging because of data gaps in eddy-covariance (EC)-based in situ flux tower observations. In this study, we obtain the latent heat flux (LE) from an EC station and perform gap filling using two deep learning methods (two-dimensional convolutional neural network (CNN) and long short-term memory (LSTM) neural networks) and two machine learning (ML) models (support vector machine (SVM), and random forest (RF)), and we investigate their accuracies and uncertainties. The average model performance based on ~25 input and hysteresis combinations show that the mean absolute error is in an acceptable range (34.9 to 38.5 Wm−2), which indicates a marginal difference among the performances of the four models. In fact, the model performance is ranked in the following order: SVM > CNN > RF > LSTM. We conduct a robust analysis of variance and post-hoc tests, which yielded statistically insignificant results (p-value ranging from 0.28 to 0.76). This indicates that the distribution of means is equal within groups and among pairs, thereby implying similar performances among the four models. The time-series analysis and Taylor diagram indicate that the improved two-dimensional CNN captures the temporal trend of LE the best, i.e., with a Pearson’s correlation of >0.87 and a normalized standard deviation of ~0.86, which are similar to those of in situ datasets, thereby demonstrating its superiority over other models. The factor elimination analysis reveals that the CNN performs better when specific meteorological factors are removed from the training stage. Additionally, a strong coupling between the hysteresis time factor and the accuracy of the ML models is observed.

A Review of Reconstructing Remotely Sensed Land Surface Temperature under Cloudy Conditions

Land surface temperature (LST) is an important environmental parameter in climate change, urban heat islands, drought, public health, and other fields. Thermal infrared (TIR) remote sensing is the main method used to obtain LST information over large spatial scales. However, cloud cover results in many data gaps in remotely sensed LST datasets, greatly limiting their practical applications. Many studies have sought to fill these data gaps and reconstruct cloud-free LST datasets over the last few decades. This paper reviews the progress of LST reconstruction research. A bibliometric analysis is conducted to provide a brief overview of the papers published in this field. The existing reconstruction algorithms can be grouped into five categories: spatial gap-filling methods, temporal gap-filling methods, spatiotemporal gap-filling methods, multi-source fusion-based gap-filling methods, and surface energy balance-based gap-filling methods. The principles, advantages, and limitations of these methods are described and discussed. The applications of these methods are also outlined. In addition, the validation of filled LST values’ cloudy pixels is an important concern in LST reconstruction. The different validation methods applied for reconstructed LST datasets are also reviewed herein. Finally, prospects for future developments in LST reconstruction are provided.