The irrigated agriculture of Southern Italy, managed by the reclamation consortia, generally lacks the necessary information and tools which can allow to improve water management and achieve the necessary sustainability of the irrigation systems. In this context, the objective of the study was to provide an effective tool for monitoring tree crop water needs at the irrigation district level. The study was carried out in Eastern-Sicily (Italy) during the years 2019–2020. An object-based classification using the Random Forest (RF) algorithm was applied to map the main tree crops of the area. The RF model was built using a temporal stack of green (B2), red (B3), and near-infrared (B8) bands of 24 selected Sentinel-2 images and 503 ground-truth sample points. The accuracy was assessed by determining the out-of-bag (OOB) error and kappa coefficient. The irrigation water requirements (IWR) were determined with the IdrAgra model (www.idragra.unimi.it) over the identified tree crop areas using the dual crop parameters provided by the FAO-56 paper and validated at 5 reference farms. The obtained IWR at the district scale were compared with the irrigation volumes distributed by the reclamation consortium. The results show the potential of hydrological simulation models coupled with remote sensing-based land use classification techniques for improving water manage-ment of tree crops and increasing the sustainability of irrigated agriculture under semi-arid conditions.

Monitoring and Predicting Irrigation Requirements of Tree Crops in Eastern Sicily as a Tool for Sustainability

Longo Minnolo G.;Vanella D.;Consoli S.
2023-01-01

Abstract

The irrigated agriculture of Southern Italy, managed by the reclamation consortia, generally lacks the necessary information and tools which can allow to improve water management and achieve the necessary sustainability of the irrigation systems. In this context, the objective of the study was to provide an effective tool for monitoring tree crop water needs at the irrigation district level. The study was carried out in Eastern-Sicily (Italy) during the years 2019–2020. An object-based classification using the Random Forest (RF) algorithm was applied to map the main tree crops of the area. The RF model was built using a temporal stack of green (B2), red (B3), and near-infrared (B8) bands of 24 selected Sentinel-2 images and 503 ground-truth sample points. The accuracy was assessed by determining the out-of-bag (OOB) error and kappa coefficient. The irrigation water requirements (IWR) were determined with the IdrAgra model (www.idragra.unimi.it) over the identified tree crop areas using the dual crop parameters provided by the FAO-56 paper and validated at 5 reference farms. The obtained IWR at the district scale were compared with the irrigation volumes distributed by the reclamation consortium. The results show the potential of hydrological simulation models coupled with remote sensing-based land use classification techniques for improving water manage-ment of tree crops and increasing the sustainability of irrigated agriculture under semi-arid conditions.
2023
978-3-031-30328-9
978-3-031-30329-6
Random Forest
Water needs
FAO-56
GIS
Sustainable irrigation management
File in questo prodotto:
File Dimensione Formato  
Pappalardo et al., 2023.pdf

solo gestori archivio

Descrizione: Conference paper
Tipologia: Documento in Post-print
Licenza: NON PUBBLICO - Accesso privato/ristretto
Dimensione 281.41 kB
Formato Adobe PDF
281.41 kB Adobe PDF   Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.11769/580209
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? ND
social impact