{"id":66285,"date":"2024-08-27T09:46:41","date_gmt":"2024-08-27T07:46:41","guid":{"rendered":"https:\/\/www.unitus.it\/?page_id=66285"},"modified":"2024-08-27T09:49:22","modified_gmt":"2024-08-27T07:49:22","slug":"tehra","status":"publish","type":"page","link":"https:\/\/www.unitus.it\/en\/research\/progetti-di-ricerca\/tehra\/","title":{"rendered":"TEHRA"},"content":{"rendered":"

\"\"<\/p>\n

Topsoil properties Estimation from Hyperspectral Remote sensing for Agriculture<\/strong><\/h3>\n

 <\/p>\n

\"\"<\/p>\n

TEHRA <\/strong><\/p>\n

\u00a0<\/strong><\/p>\n

ABSTRACT <\/strong><\/p>\n

The project THERA aims at developing methods and algorithms for the generation of PRISMA products for monitoring and mapping soil properties of agronomic and environmental interest, in order to foster the adoption of more sustainable and climate-smart precision farming strategies and policies.<\/p>\n

The topsoil properties targeted by the project, include primary variables directly affecting specific spectral features, such as: soil texture (clay, sand and silt percentage), soil organic carbon (SOC), soil moisture, iron oxides, salinity and calcium carbonates. Additionally, the project also targets some secondary variables, correlated with the primary variables: potassium (K), total nitrogen (N), Cation Exchange Capacity (CEC) and hydraulic properties such as field capacity and available water content (AWC).<\/p>\n

The methodologies of this project have been conceived to confront the main issues that hinder the mainstreaming of satellite-based imaging spectroscopy in precision agriculture and agricultural soil monitoring for policy related objectives.<\/p>\n

The development of topsoil properties estimation algorithms are carried out in the TEHRA project using state-of-the-art approaches covering three group of methods: 1) spectral features analysis; 2) chemometric and multivariate and regression methods; 3) machine learning approaches. The latter especially focuses on the use of deep learning methods such as convolutional neural networks.<\/p>\n

The issue of generalisation of the calibration of multivariate and machine learning regression models, from local to larger application domains, are dealt with by investigating methodologies exploiting large national\/continental soil spectral libraries, as well as the use of internal standards and transfer methods from the lab to the field.<\/p>\n

 <\/p>\n

Identification Code: ASI Contract No. 2022-6-U. 0. CUP no. F83C22000160005<\/p>\n

Total project cost: \u20ac 428.001,81 \u20ac.<\/p>\n

Total cost DAFNE: 157.999 \u20ac (cofin al 50% : finanziamento ASI \u20ac 77.999,99 \u20ac)<\/p>\n

DAFNE Role: Lead Partner
\nScientific Supervisor DAFNE: Prof Raffaele Casa<\/p>\n

Email: \u00a0 rcasa@unitus.it<\/a><\/p>\n

Start date: 05\/05\/2022.<\/p>\n

End date: 05\/11\/2024<\/p>\n

\u00a0<\/strong><\/p>\n

LINK <\/strong><\/p>\n

https:\/\/www.asi.it\/wp-content\/uploads\/2023\/12\/Casa.pdf<\/a><\/p>\n

 <\/p>\n

 <\/p>\n

PARTNERSHIP<\/p>\n

The Institutes participating to the TEHRA project are 2 Italian public bodies and a private company:<\/p>\n