Soil water content assessment: Critical issues concerning the operational application of the triangle method
- Autori: Maltese, A.; Capodici, F.; Ciraolo, G.; La Loggia, G.
- Anno di pubblicazione: 2015
- Tipologia: Articolo in rivista (Articolo in rivista)
- OA Link: http://hdl.handle.net/10447/210137
Knowledge of soil water content plays a key role in water management efforts to improve irrigation efficiency. Among the indirect estimation methods of soil water content via Earth Observation data is the triangle method, used to analyze optical and thermal features because these are primarily controlled by water content within the near-surface evaporation layer and root zone in bare and vegetated soils. Although the soil-vegetation-atmosphere transfer theory describes the ongoing processes, theoretical models reveal limits for operational use. When applying simplified empirical formulations, meteorological forcing could be replaced with alternative variables when the above-canopy temperature is unknown, to mitigate the effects of calibration inaccuracies or to account for the temporal admittance of the soil. However, if applied over a limited area, a characterization of both dry and wet edges could not be properly achieved; thus, a multi-temporal analysis can be exploited to include outer extremes in soil water content. A diachronic empirical approach introduces the need to assume a constancy of other meteorological forcing variables that control thermal features. Airborne images were acquired on a Sicilian vineyard during most of an entire irrigation period (fruit-set to ripening stages, vintage 2008), during which in situ soil water content was measured to set up the triangle method. Within this framework, we tested the triangle method by employing alternative thermal forcing. The results were inaccurate when air temperature at airborne acquisition was employed. Sonic and aerodynamic air temperatures confirmed and partially explained the limits of simultaneous meteorological forcing, and the use of proxy variables improved model accuracy. The analysis indicates that high spatial resolution does not necessarily imply higher accuracies.