Irrigation volume assessment via Remote Sensing and Cloud Engine tools

The aim of this project is to design and develop an irrigation volume assessment via remote sensing and cloud engine tools. This product objective is the estimation of soil moisture on agricultural fields in the North of Italy. The project was carried out not only to control the irrigation of this area, but also to increase the knowledge of the crops and the vegetation state to help to obtain the best possible agricultural products. The project was developed under the framework of the space technologies. Since 2010, with the launch of the Soil Moisture and Ocean Salinity (SMOS) Satellite conducted by the European Space Agency (ESA), or with the Soil Moisture Active Passive (SMAP) Satellite launch by the NASA in 2015, measuring soil moisture remotely became possible. Despite their huge advantages, some limitations made the data from these satellites not useful for agricultural applications, especially their low spatial resolution (30km-50km). In 2014, due to the progress the space sector has been through the last years, the Copernicus Program was created as a work collaboration between the ESA and the European Union (EU). The objective of this program is to obtain exact and continuous information in order to improve the management and conservation of the environment. Furthermore, it seeks to better understand and mitigate the effect of the climate change. Finally, it looks to ensure the civil security. By the hand of this program, a constellation of satellites, the Sentinel Satellites, have been created. These satellites are characterized by covering different types of Earth themes, i.e. oceans, land, atmosphere, etc. and to have a really high spatial resolution, around the dozen of meters. Unfortunately, among all the variables measured by these satellites soil moisture is not one of them. Therefore, the scope of this project was to estimate the soil moisture values from the raw data of the satellites. The soil moisture estimation task was carried out by transforming the backscatter values measured by the Sentinel 1 satellite into the desire variable. This transformation is developed by applying a mathematical model that relates backscatter with soil moisture. There are three big categories of these methods: Neural Networks (NN) approach, Water Cloud Models (WCM) and Change Detection (CD) methods. Due to the data available to develop the task the most suitable category for this project was CD methods. Among all the existent CD methods the one chosen was the Copernicus Global Land Service method, due to the programming limitations of the cloud engine. A cloud engine is understood as an online platform where both, raw data and already developed products, can be obtained. Moreover, one of the main characteristics of these tools is the possibility to work online with the raw data in order to create a specific product that fulfills all the proposed requirements. This project was developed in the Google Earth Engine (GEE) Code Editor platform, which is an application developed by Google. Its aim is the scientific analysis at a petabyte scale and the visualization of geospatial data obtained through remote sensing. It seeks for public benefit, as well as for commercial users benefit and Administrational. The results of the project were the correct collecting of soil moisture estimations. The method applied was validated comparing the results obtained with already validated data, with a RMSD of 8.5% vol.