Читать книгу Climate Impacts on Sustainable Natural Resource Management - Группа авторов - Страница 42

For decades remote sensing has resulted in the collection of huge volumes of datasets. The management and analyzing of these voluminous datasets cannot practically be achieved using standard software packages and general computing systems. To address this challenge, Google has developed the first cloud computing platform of its kind, called Google Earth Engine (GEE), for effectively accessing and processing these datasets. GEE facilitates big geo‐data processing at country, continental, or world level and provides datasets for long periods (Amani et al. 2020). All the publicly available remote sensing data from multiple satellites, such as the Landsat series, Moderate Resolution Imaging Spectrometer (MODIS), Sentinel series, National Oceanographic and Atmospheric Administration Advanced very high‐resolution radiometer (NOAA AVHRR), Advanced Land Observing Satellite (ALOS), along with other gridded datasets is used. The complete list of datasets is available on the GEE webpage (https://earthengine.google.com/datasets) (Kumar and Mutanga 2018).

GEE connects with Google's computational infrastructure and provides users with access to the dataset in the backend using a web interface. Javascript and Python are the two scripting languages used to retrieve the data, process it, and analyze it. GEE has parallel processing and fast computation features to deal with a huge volume of data processing challenges effectively. In this cloud computing platform, the users do not download the dataset. Instead, all the processing is completed at the Google server‐side and the only final product can be downloaded. No additional software other than a web browser is required to perform analysis on GEE. Users have the flexibility to utilize software like various Python Integrated Development Environments (IDEs) or a simple browser interface. GEE also provides many built‐in algorithms for cloud correction, classification, and matrix operations as well. These algorithms are easy to use tools to analyze data at a planetary scale and can be used as a building block for developing user‐defined algorithms by scientists with less effort. Some of the wide range of GEE applications, including those for NRM, are shown in Figure 2.7. Its application for image classification is shown in Figure 2.6.