This is the Google scholar page of the paper. On the map pop up (when clicking a paper point on the map, the popped-up info box) and on the data table, which is at the bottom of the web app, there is a clickable URL that will direct you to the paper’s Google Scholar page. This is where readers can find new papers citing each paper, so new paperas can be easily tracked. Note that if a paper does not have a Google Scholar page, we provide the ResearchGate link of the paper, if available.

Not every paper seems to have a Google Scholar page.However, if a paper has at least one citation, it will have a Google Scholar Cited by URL to allow our users/readers to easily track new publications citing the paper.

This field is for what application a paper focused on (e.g., vegetation, wetland, water mapping).

Method: The authors used novel CV, ML, or DL methods. Note that we determined this based on the CV, ML, DL part of the analysis, not on pre-processing methods even if they were novel. So this value is used to represent a paper that moves forward the integration of GEE with AI and its branches, CV/ML/DL.
Application: This could look like the authors using ready-to-use models on GEE with their own data or data they found on GEE (“plug-and-play”).
(Note: this field is one of the three fields used to build the methods hierarchical sunburst chart)

The metrics the authors used to evaluate their model(s) and prediction(s).

The CV algorithms and/or ML/DL models the authors use in their paper.

The location where the study focused on. If there is more than one place analyzed, we list the countries separated by a comma. If it is at global scale, we provide the value “global”.

The cloud computing platform the authors use, if any. Note that not all authors use GEE for processing or analyzing data. Some authors simply used the platform for finding and downloading data, in which case this column should be empty (i.e., blank). Also, GEE is not the only possible value, though it is the most common. For example, one set of authors combined GEE and Google Cloud AI, where a few used Google Colab and Google Drive. Note that, if authors provide hardware specifications (e.g., GPU, RAM, CPU, and runtimes) about the cloud platform they used, we also report the information here.

This column is for specifying what hardware (e.g., GPU, CPU, RAM, hard disk, runtimes) authors used if they completed part of their analysis offline. However, the majority of studies do not report hardware specifications, runtimes, etc. so this column will be listed as “NS”, which refers to “not specified”. For those only used cloud platforms and no local computers involved, the entry will be “NA” for “not applicable”.

This column is useful for identifying which software, including libraries (if specified), authors use to complete their analysis. Software listed here are for both online and/or offline analyses. Examples are: ArcGIS, R, Python, MATLAB, eCognition, etc. Some authors do not specify which software they use to complete their analysis, in which case “NS” or “offline/NS” are listed, which mean “not specified” and “offline - not specified” respectively. Note: if a publication only> used Google Earth Engine or its APIs, this is not listed under this “Other Software” column, but instead given a value of “NA” for “not applicable”. Instead, it is included in the “Computed on Cloud” column because GEE is both a cloud computing and online software platform. Additionally, not all authors specify whether they use JavaScript, Python, or R when working specifically with GEE (and not other software packages within those programming languages), so it would be difficult to specify in any useful detail how authors were using GEE software for this column.

The types of RS data or RS-derived data products the authors used in their analysis, either for direct analysis and processing or for validation purposes. (Note: This does not include whether or not the authors used field data or some other type of non-RS data.)

Whether or not the authors used machine learning (ML), deep learning (DL), and/or computer vision (CV). If a paper uses more than one method, we list them and separate them by a comma.
(Note: this field is one of the three fields used to build the methods hierarchical sunburst chart)

Classification, segmentation, regression, or some combination of these choices, separated by a comma.
(Note: this field is one of the three fields used to build the methods hierarchical sunburst chart)