Loading...
Publication
Performance of ESA-CCI-SM Satellite product for estimating soil water deficit: a case study in NE Portugal
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 2 MB | Adobe PDF |
Advisor(s)
Abstract(s)
The Mediterranean climate is characterised by hot and dry summers and water deficit in soils. Monitoring soil water deficit is of crucial importance in Mediterranean cropland, allowing a more cost-effective soil and water management. Satellite-borne data sources provide high temporal resolution information but usually lack the spatial resolution to adequately account for soil variability, namely on the physical-chemical properties that affect soil water storage capacity. As a case study located in Bragança, NE Portugal, the present work aimed at exploring the use of ESA-CCI-SM satellite product to estimate water deficit for a wide range of soil water storage capacities. A satellite-borne soil moisture data series (ESA-CCI-SM product) was clipped for Bragança area (25 x 25 km pixel). Temperature and precipitation (P) records of a weather station located in Bragança were used to compute monthly reference evapotranspiration (ET0, Thornthwaite) and the soil water balance (SWB, Thornthwaite-Mather) along the same period (2003-2016, 168 months). SWB were computed for soil water storages (Smax) representing the Smax range of regional soils (25, 50, 75, 100, 150 mm). Following earlier work by the authors, monthly soil water deficit series (SWD), obtained for the 5 Smax, were analysed and compared with the SM monthly averages series, using basic statistical methods. Average annual P and ET0 in the study period were 750 and 711 mm, respectively. In all years, SWD occurred from June to September, with peaks in July and August (ca. 110 mm each), but it was also recorded in February and in November (1 out of 14 years). Annual SWD decreased linearly with Smax increase (R²=0.994), from 400 mm (Smax 25 mm) to 290 mm (Smax 150 mm). The number of SWD months also decreased in the same range of Smax, from 77 to 74 out of 168. Linear regression models relating SWD to SM monthly data for each Smax were derived and applied to estimate SWD in dry months. Model performance declined with Smax increase, from Smax 25 mm (R2=0.928) to Smax 150 mm (R2=0.873), while SEE remained similar along the Smax range (23 mm). Model parameters are well correlated (R2>0.95) with Smax, allowing the integration in a single model the two variables (SM and Smax). Dry months (SWD>0) were discriminated from non-dry (SWD=0) by SM, respectively lower or higher than 0.21 m3 m-3, with 91% (Smax 25 mm) to 94% (Smax 150 mm) matching frequency. Median SM in the whole Smax range was 0.14 and 0.25 m3 m-3, for dry and non-dry months, respectively. Performance reached encourages the application of the regression models derived for estimating SWD from SM, still with a considerable uncertainty due to the high SEE obtained. As the range of Smax tested was quite large, the regression models can be applied to deliver remote-based estimates of SWD accounting for the variability of soil water storage capacities found in NE Portugal.
Description
Keywords
Soil water deficit
Pedagogical Context
Citation
Figueiredo, Tomás d'Aquino; Royer, Ana Caroline; Hernandez Hernandez, Zulimar (2025). Performance of ESA-CCI-SM Satellite product for estimating soil water deficit: a case study in NE Portugal. In Advancing Soil Knowledge for a Sustainable Future - EUROSOIL 2025: Book of Abstracts. Sevilla. ISBN 978-84-09-75471-7
Publisher
Difital.CSIC