High-resolution satellite products improve hydrological modeling in northern Italy

Abstract

Abstract. Satellite-based Earth observations (EO) are an accurate and reliable data source for atmospheric and environmental science. Their increasing spatial and temporal resolutions, as well as the seamless availability over ungauged regions, make them appealing for hydrological modeling. This work shows recent advances in the use of high-resolution satellite-based EO data in hydrological modeling. In a set of six experiments, the distributed hydrological model Continuum is set up for the Po River basin (Italy) and forced, in turn, by satellite precipitation and evaporation, while satellite-derived soil moisture (SM) and snow depths are ingested into the model structure through a data-assimilation scheme. Further, satellite-based estimates of precipitation, evaporation, and river discharge are used for hydrological model calibration, and results are compared with those based on ground observations. Despite the high density of conventional ground measurements and the strong human influence in the focus region, all satellite products show strong potential for operational hydrological applications, with skillful estimates of river discharge throughout the model domain. Satellite-based evaporation and snow depths marginally improve (by 2 % and 4 %) the mean Kling–Gupta efficiency (KGE) at 27 river gauges, compared to a baseline simulation (KGEmean= 0.51) forced by high-quality conventional data. Precipitation has the largest impact on the model output, though the satellite data on average shows poorer skills compared to conventional data. Interestingly, a model calibration heavily relying on satellite data, as opposed to conventional data, provides a skillful reconstruction of river discharges, paving the way to fully satellite-driven hydrological applications.

IT-SNOW: a snow reanalysis for Italy blending modeling, in-situ data, and satellite observations (2010–2021)

“…In terms of epistemic uncertainty, the S3M model relies on an enhanced temperature-index approach that was calibrated in Aosta valley and then extensively evaluated elsewhere in Italy (both in this paper and in other projects, such as Alfieri et al, 2022). While relying on the same parameters across the whole country could introduce some additional uncertainty at local scale, we highlight that results in this paper show a credible reconstruction of melt dynamics even in areas with only occasional assimilation (Figure 7), while Bouamri et al (2018) reported encouraging results when transferring model parameters of the same enhanced temperature-index approach to uncalibrated sites.…”

Parameter transferability of a distributed hydrological model to droughts

“…The model setup used here consisted of six modules (namely, the snow, vegetation, energy balance, soil, groundwater, and surface water modules) to simulate snow dynamics, vegetation interception, energy fluxes, evaporation from canopy layer, evapotranspiration, soil moisture and groundwater dynamics, and streamflow generation (Figure S1). Further, we simulated major lakes and dams in the region (Alfieri et al, 2022). We refer the reader to Silvestro et al (2013) for a description of the model, Silvestro et al (2015) for specifics on the snow module, and Silvestro et al (2021) for specifics on the surface flow routing scheme.…”

“…In this work, we run Continuum at a 0.009°(around 1 km) spatial resolution and 1 hour temporal resolution (Alfieri et al, 2022) over the hydrological years (h.y.) 2009 -2022, with the first h.y.…”

“…In this work, we used the model setup and the required datasets, regarding topography, soil properties, land cover, dams, lakes, and glacier cover, presented in Alfieri et al (2022). We summarized the input data in Table S1 and we refer the reader to Alfieri et al (2022) for a detailed description of them.…”

“…For precipitation we used interpolated maps from groundand radar-based P data from the Italian Civil Protection Department (in the following DPC), computed with the Modified Conditional Merging algorithm (Bruno et al, 2021). For other meteorological variables we used interpolated maps based on ground-based data provided by DPC (see Alfieri et al (2022) for details).…”