Author(s): Mohammad Zeynoddin; Hossein Bonakdari; Silvio Jose Gumiere; Jean Caron; Alain N. Rousseau

Linked Author(s): Mohammad Zeynoddin, Hossein Bonakdari

Keywords: FLDAS; Metaheuristic; State-space; Google earth engine; Soil moisture; Soil temperature; Remote sensing

Abstract: Over the past two decades, the development and invention of new methods of collecting and disseminating information have tremendously facilitated access to data, especially remote sensing data with unprecedented temporal, spatial and spectral resolutions. This possibility has been empowered by the development of cloud application (app) programming interfaces (CAPIs) for storing, analyzing and disseminating large volumes of data. Google Earth Engine (GEE) is one of the CAPIs; providing access to big data available through huge geo-environmental catalog. Given these achievements, we developed a web-based online app to allow public access to near real-time satellite and land-surface based surface soil moisture (SSM) and temperature (SST) data as easy as clicking an icon or typing coordinates. The developed web-based online app - soilparam 1.0 - is accessible through the following link: https://zemoh.users.earthengine.app/view/soilparam The first version of soilparam provides access to the NASA-USDA Enhanced SMAP Global soil moisture dataset and the first layer of the Famine Early Warning Systems Network Land Data Assimilation System soil temperature dataset and converts the collection of images into time series. This app has solved the problem of handling big data often found in remote-sensing studies and provides users with a very friendly and secure environment. Following the introduction of the app, the process of obtaining the time series of the aforementioned variables is described for any location in the world, along with pseudo-codes. In this study, the SSM and SST datasets of a region in Quebec, Canada, were downloaded by the app to perform geo-spatial and time series analysis. Moreover, these datasets are modeled with a metaheuristic-space-state model, suitable for seasonal variables to obtain future projections. The results demonstrate that the evolutionary technique, integrated with the advanced state-space models, can forecast SST data with an acceptable accuracy for both dynamic and periodical forecasts. Adding more datasets to the app with different resolutions, parameters and depths, is planned.

DOI: https://doi.org/10.3850/IAHR-39WC252171192022808

Year: 2022