Remote Sensing (RS) systems have been collecting massive volumes of datasets for decades, managing and analyzing of which are not practical using common software packages and desktop computing resources. In this regard, Google has developed a cloud computing platform, called Google Earth Engine (GEE), to effectively address the challenges of big data analysis. In particular, this platform facilitates processing big geo data over large areas and monitoring the environment for long periods of time. Although this platform was launched in 2010 and has proved its high potential for different applications, it has not been fully investigated and utilized for RS applications until recent years. Therefore, this study aims to comprehensively explore different aspects of the GEE platform, including its datasets, functions, advantages/limitations, and various applications. For this purpose, 450 journal articles published in 150 journals between January 2010 and May 2020 were studied. It was observed that Landsat and Sentinel datasets were extensively utilized by GEE users. Moreover, supervised machine learning algorithms, such as Random Forest (RF), were more widely applied to image classification tasks. GEE has also been employed in a broad range of applications, such as Land Cover/land Use (LCLU) classification, hydrology, urban planning, natural disaster, climate analyses, and image processing. It was generally observed that the number of GEE publications has significantly increased during the past few years, and it is expected that GEE will be utilized by more users from different fields to resolve their big data processing challenges.

Although wetlands provide valuable services to humans and the environment and cover a large portion of Canada, there is currently no Canada-wide wetland inventory based on the specifications defined by the Canadian Wetland Classification System (CWCS). The most practical approach for creating the Canadian Wetland Inventory (CWI) is to develop a remote sensing method feasible for large areas with the potential to be updated within certain time intervals to monitor dynamic wetland landscapes. Thus, this study aimed to create the first Canada-wide wetland inventory using Landsat-8 imagery and innovative image processing techniques available within Google Earth Engine (GEE). For this purpose, a large amount of field samples and approximately 30,000 Landsat-8 surface reflectance images were initially processed using several advanced algorithms within GEE. Then, the random forest (RF) algorithm was applied to classify the entire country. The final step was an original CWI map considering the five wetland classes defined by the CWCS (i.e., bog, fen, marsh, swamp, and shallow water) and providing updated and comprehensive information regarding the location and spatial extent of wetlands in Canada. The map had reasonable accuracy in terms of both visual and statistical analyses considering the large area of country that was classified (9.985 million km²). The overall classification accuracy and the average producer and user accuracies for wetland classes exclusively were 71%, 66%, and 63%, respectively. Additionally, based on the final classification map, it was estimated that 36% of Canada is covered by wetlands.

Wetlands are an important natural resource that requires monitoring. A key step in environmental monitoring is to map the locations and characteristics of the resource to better enable assessment of change over time. Synthetic Aperture Radar (SAR) systems are helpful in this way for wetland resources because their data can be used to map and monitor changes in surface water extent, saturated soils, flooded vegetation, and changes in wetland vegetation cover. We review a few techniques to demonstrate SAR capabilities for wetland monitoring, including the commonly used method of grey-level thresholding for mapping surface water and highlighting changes in extent, and approaches for polarimetric decompositions to map flooded vegetation and changes from one class of land cover to another. We use the Curvelet-based change detection and the Wishart-Chernoff Distance approaches to show how they substantially improve mapping of flooded vegetation and flagging areas of change, respectively. We recommend that the increasing availability SAR data and the proven ability of these data to map various components of wetlands mean SAR should be considered as a critical component of a wetland monitoring system.

Advances in machine learning have changed many fields of study and it has also drawn attention in a variety of remote sensing applications. In particular, deep convolutional neural networks (CNNs) have proven very useful in fields such as image recognition; however, the use of CNNs in large-scale remote sensing landcover classifications still needs further investigation. We set out to test CNN-based landcover classification against a more conventional XGBoost shallow learning algorithm for mapping a notoriously difficult group of landcover classes, wetland class as defined by the Canadian Wetland Classification System. We developed two wetland inventory style products for a large (397,958 km²) area in the Boreal Forest region of Alberta, Canada, using Sentinel-1, Sentinel-2, and ALOS DEM data acquired in Google Earth Engine. We then tested the accuracy of these two products against three validation data sets (two photo-interpreted and one field). The CNN-generated wetland product proved to be more accurate than the shallow learning XGBoost wetland product by 5%. The overall accuracy of the CNN product was 80.2% with a mean F1-score of 0.58. We believe that CNNs are better able to capture natural complexities within wetland classes, and thus may be very useful for complex landcover classifications. Overall, this CNN framework shows great promise for generating large-scale wetland inventory data and may prove useful for other landcover mapping applications.

Despite providing vital ecosystem services, wetlands are increasingly threatened across the globe by both anthropogenic activities and natural processes. Synthetic aperture radar (SAR) has emerged as a promising tool for rapid and accurate monitoring of wetland extent and type. By acquiring information on the roughness and moisture content of the surface, SAR offers unique potential for wetland monitoring. However, there are still challenges in applying SAR for mapping complex wetland environments. The backscattering similarity of different wetland classes is one of the challenges. Choosing the appropriate SAR specifications (incidence angle, frequency and polarization), based on the wetland type, is also a subject of debate and should be investigated more thoroughly. The geometric distortion of SAR imagery and loss of coherency are other remaining challenges in applying SAR and its processing techniques for wetland studies. Hence, this study provides a systematic meta-analysis based on compilation and analysis of indexed research studies that used SAR for wetland monitoring. This meta-analysis reviewed 172 papers and documented an upward trend in usage of SAR data, increasing usage of multi-sensor data, increasing integration of C- and L- bands over other configurations and higher classification accuracy with multi-frequency and multi-polarized SAR data. The highest number of wetland research studies using SAR data came from the USA, Canada and China. This meta-analysis highlighted the current challenges and solutions for wetland monitoring using SAR sensors.

While deep learning models have been extensively applied to land-use land-cover (LULC) problems, it is still a relatively new and emerging topic for separating and classifying wetland types. On the other hand, ensemble learning has demonstrated promising results in improving and boosting classification accuracy. Accordingly, this study aims to develop a classification system for mapping complex wetland areas by incorporating deep ensemble learning and satellite datasets. To this end, time series of Sentinel-1 dual-polarized Synthetic Aperture Radar (SAR) dataset, alongside Sentinel-2 multispectral imagery (MSI), are used as input data to the model. In order to increase the diversity of the extracted features, the proposed model, herein called WetNet, consists of three different submodels, comprising several recurrent and convolutional layers. Furthermore, multiple ensembling sections are added to different stages of the model to increase the transferability of the model (to other areas) and the reliability of the final results. WetNet is evaluated in a complex wetland area located in Newfoundland, Canada. Experimental results indicate that WetNet outperforms the state-of-the-art deep models (e.g., InceptionResnetV2, InceptionV3, and DenseNet121) in terms of both the classification accuracy and processing time. This makes WetNet an efficient model for large-scale wetland mapping application. The python code of the proposed WetNet model is available at the following link for the sake of reproducibility: https://colab.research.google.com/drive/1pvMOd3_tFYaMYGyHNfxqDxOiwF78lKgN?usp=sharing