Abstract
Droughts are a recurring climatic hazard with significant ecological and socioeconomic consequences. Their frequency and severity have intensified under global warming, driven by rising evapotranspiration, declining soil moisture, and intensified hydrological cycles. Even historically water-rich regions like northern Italy are increasingly affected. To monitor and understand drought events and their impacts, drought indices are essential. The Standardized Precipitation Evaporation Index (SPEI) is a widely used drought indicator representing the deviations in water balance, i.e., the difference between precipitation and potential evapotranspiration (PET), thus allowing the temperature effects to be considered. As a standardized indicator, it facilitates the result comparison across different environments and climate conditions and enables analysis across different time scales, from short-term agricultural droughts to longer-term hydrological or physiological droughts. Drought indices are generally defined at monthly resolution. However, for specific applications, such as the comparison of meteorological conditions with soil moisture dynamics or with the response of vegetation, daily indices might be preferable since they enable a more detailed description of the onset and end time of deficit conditions and of short-term variations. This study assesses the methods for the calculation of SPEI at daily scale and evaluates its suitability with respect to monthly SPEI to capture drought episodes and represent the spatio-temporal relationship with vegetation conditions in forest ecosystems. The analysis was conducted in Mazia valley, an inner-alpine dry valley in the eastern Italian Alps (Trentino-South Tyrol), and SPEI values were derived from a daily gridded observational dataset (250 m resolution) of temperature and precipitation covering the study area from 1980 to 2024. For the calculation of daily SPEI, several probability distributions and fitting strategies for water balance values (precipitation minus PET) were tested. The use of a moving window centred on each day was found to be the best choice reducing result variability and reflecting seasonality. Daily SPEI time series were calculated at various aggregation scales (15 days to 6 months) and drought events were defined at each grid-cell level as periods of at least 30 consecutive days with negative SPEI values, starting when the index dropped below a certain threshold depending on the drought class considered (moderate to extreme). Comparing daily and monthly SPEI in terms of drought detection confirmed that the higher temporal resolution better captures the short-term yet intense droughts and the inter-monthly variability of water balance. When daily SPEI was compared to vegetation water stress indicators, including remotely sensed spectral indices and tree-level measurements collected for alpine forests in Mazia valley, they were found to allow for a better comparability with physiological drought conditions. Preliminary analyses provide insights on the use of daily SPEI for high-resolution monitoring of drought and understanding of drought dynamics and impacts.