Naslov
Hyperspectral remote sensing of drought in vineyards

Autori
Žibrat, Uroš ; Knapič, Matej ; Zovko, Monika

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni

Izvornik
7th Slovenian Symposium on Plant Biology with International Participation, Biotechnical Faculty, University of Ljubljana, September 17-18, 2018, Ljubljana, Slovenia / Baebler, Špela ; Dermastia, Marina ; Grebenc, Tine ; Tomaž, Špela ; Železnik, Peter - Ljubljana : Silva Slovenica Publishing Centre, Slovenian Foresty Institute, 2018, 18-18

ISBN
978-961-6993-44-9

Skup
7th Slovenian Symposium on Plant Biology with international participation

Mjesto i datum
Ljubljana, Slovenija, 17.09.2018. - 18.09.2018

Vrsta sudjelovanja
Poster

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
remote sensing, karst, soil, irrigation

Sažetak
In the Mediterranean many parts of karst surfaces have been modified to make them suitable for agricultural use, mostly grapevine and olive production. In karst landscapes rain water percolates rapidly downward through the soil layers, making them highly fragile and vulnerable to anthropogenic as well as climatic stresses. Moreover, excessively dry topsoils can cause water run-off and significant evaporative loss of water, making lack of moisture an important risk indicator in vine production. Water measurements in karst soils are challenging, due to skeletal structure and great within-field variability. Under such conditions, irrigation scheduling should be mainly based on crop water status. Current standard procedures include measurements of individual plants, leading to extensive and time consuming field work, which is subject to measurement and sampling errors. We assessed the applicability of hyperspectral remote sensing for grapevine water status assessment. The research was performed in an experimental vineyard grown in artificially transformed kart terrain in Croatia, near Šibenik. The experimental design included four water treatments in three replicates: (1) fully watered (100% ETc), (2) 75% ETc, (3) 50% ETc, and (4) non-watered. Hyperspectral images were taken using two cameras (VNIR and SWIR), covering wavelengths fom 400 to 2500 nm. Data were analysed using Partial least squares discriminant analysis and Support vector machines. We achieved a classification success of 70 - 80% for determining watering status. The 75 and 100% groups could not be reliably distinguished between each other.

Izvorni jezik
Engleski

Znanstvena područja
Poljoprivreda (agronomija)

POVEZANOST RADA