MAPPING EVAPOTRANSPIRATION ON VINEYARDS: THE SENTINEL-2 POTENTIALITY

Abstract

stimation of actual evapotranspiration in Sicilian vineyards, is an emerging issue since these agricultural systems. Indeed unlike other agricultural species (Vitis vinifera L. ) are generally cultivated under mild water stress, in order to enhance quality (Guadillère et al. , 2002. This has significant impacts on the management of the scarce water resources of the region. The choice of the most appropriate methodology for assessing water use in these systems is still an issue of debating, due to the complexity of canopy and root systems and for their high spatial fragmentation. In vineyards, quality and quantity of the final product are dependent on the controlled stress conditions to be set trough irrigation. This paper reports an application of the well-known Penman-Monteith approach, applied in a distributed way, using high resolution remote sensing data to map the potential evapotranspiration (ETp ). In 2008 a series of airborne multispectral images were acquired on the "Tenute Rapitalà", a wine farm located in the northwest of Sicily. Five airborne remote sensing scenes were collected using a SKY ARROW 351 650 TC/TCNS aircraft, at a height of about 1000 m a.g.l.. The acquisitions encompassed almost a whole phenological period, between June and September 2008 (approximately one each three weeks). The platform had on board a multi-spectral camera with 3 spectral bands in the green (G, 530-570 nm), red (R, 650-690 nm) and near infrared (NIR, 767-832 nm) wavelengths, and a thermal camera with a broad band in the range 7.5-13 μ m. The nominal pixel resolution was approximately 0.7 m for VIS/NIR acquisitions, and 1.7 m for the thermal-IR data. Field data were acquired simultaneously to airborne acquisitions. The former include spectral reflectance in visible, near infrared, middle infrared (VIS, NIR, MIR) regions of the spectrum, leaf area index (LAI), soil moisture at different depths (both in row and below plants). Moreover, meteorological variables and fluxes were acquired by means of an Eddy correlation tower located within the field. The VIS-NIR bands were atmospherically corrected and calibrated in order to calculate shortwave albedo (SW ), vegetation height (hC ) and LAI, representing the distributed inputs of the Penman-Monteith equation. The paper reports a comparison between the application on the original airborne imagery and on the degraded ones in order to simulate the Sentinel-2 spatial resolution. Moreover a sensitivity analysis has been carried out on input parameters (i.e. SW , LAI ). The sensitivity was carried out to highlight the variability of ETP on the accuracy in the parameters assessment obtainable using the multispectral Sentinel-2 images, compared to the one retrieved processing high spatial resolution airborne images. The coarser spatial detail of Sentinel-2 images is probably compensated by an increase in spectral resolution, but this will be investigated in next papers.