Search

The urgency of understanding the intricate input–output relationships of the hydrologic cycle is amplified by the accelerating climate change impacts in mountain environments. This study focuses on optimising water resource management of a dammed valley in the Central Alps (Northern Italy). The research aims to integrate radar data and precipitation interpolation techniques (TIN, Copula, cumulative distribution function; CDF techniques, inverse distance weighting; IDW, thin plate spline; TPS, ordinary kriging; OK and detrended kriging; DK) into a semi‐distributed hydrologic model, by utilising hourly precipitation data from 22 rain gauges and a composite weather radar product spanning 2010–2020. Two main objectives were pursued: (i) to develop and evaluate various radar precipitation correction methods against a benchmark dataset and (ii) to calibrate and assess the performance of the GEOFrame hydrologic model forced with corrected precipitation input. Point‐based and spatial correction approaches were evaluated against ground measurements through leave‐one‐out tests. The former derives dependence functions between the biased radar series and those of the closest three rain gauges to the target point applying a triangular irregular network. The latter combines deterministic and geospatial interpolations to the rain gauge/radar residuals to derive a corrected surface by incorporating radar values as trends. Precipitation series exceeding the composite scaled score of the benchmark dataset were used as input for hydrologic modelling. The spatial method combining radar values with ordinary kriging provided the best results for both correction and modelling (hourly KGE > 0.75). The spatial approaches proved easier to apply than the point‐based methods. In addition, correcting precipitation significantly improved low‐flow simulation from negative hourly lnNSE to values greater than 0.25. As a further step, given the overall good performance of the spatial methods, they could be used operationally as an ensemble to analyse management scenarios. [ABSTRACT FROM AUTHOR]

Climate models are useful tools for analyzing historical and projecting future climate conditions. However, the model results tend to differ systematically from observations, particularly for parameters with complex spatial and temporal distributions such as precipitation. A combination of quantile mapping and generalized additive models (GAMs) is presented and proposed as a new method (Q‐GAM) for the bias correction of daily precipitation. Q‐GAM is demonstrated by using data from five European stations with different climate characteristics. For each station, the closest continental grid point of a EURO‐CORDEX climate model was selected for bias correction. A bootstrapping experiment is presented with over 1000 repetitions of randomly splitting the historical period 1981 to 2005 into a calibration and evaluation period. The results for all stations reveal that Q‐GAM is a straightforward, accurate and computationally efficient method for the bias correction of precipitation. In particular, the method improves the frequency of dry days and the total annual rainfall amount. This outcome is robust across stations with varying climate characteristics and also to the choice of calibration and evaluation periods. Similar results are also obtained for other precipitation characteristics such as the 0.9 and 0.95 quantiles. [ABSTRACT FROM AUTHOR]

The urban heat island (UHI) effect denotes that temperatures in urban areas are often higher compared to the rural background. This human‐induced phenomenon can exacerbate heat stress and adverse health impacts, adding to ongoing global warming. To examine the intensity and variability of the UHI effect, we used 40 years (1980−2019) of observational data (daily maximum and minimum temperature) from the Global Summary of the Day (GSOD), comprising about 1000 stations of varying temporal extent, spanning the Middle East and North Africa (MENA). The MENA is characterised by hot and dry summers and warms faster than other inhabited regions worldwide. The challenge in using data with diverse spatial and temporal extents is accounting for heterogeneities between each station comprising an urban–rural pair. One has to allow for differences in the distance between the pairs and, elevation, spatial–temporal changes in urbanisation as well as the distance from the coast. A new method is proposed, based on flexible statistical methods (Generalized Additive Models or GAMs), to quantify the temporal trend in the UHI effect while allowing for the aforementioned characteristics using regression splines of appropriately defined variables. A composition of high‐resolution satellite geospatial information, related to urbanisation properties and population data was utilised from the Global Human Settlement Layer database (GHSL‐SMOD), to characterise the stations in terms of their urbanisation type. These data were also used to quantify temporal changes in the extent of urbanisation relative to the surrounding areas. Results indicate consistent upward trends of the UHI effect in the MENA (up to ~0.5°C), particularly at night (daily minimum temperature) during all seasons. [ABSTRACT FROM AUTHOR]

Precipitation holds significant importance as a climate parameter in various applications, including studies on the impacts of climate change. However, its simulation or projection accuracy is low, primarily due to its high stochasticity. Specifically, climate models often overestimate the frequency of light rainy days while simultaneously underestimating the total amounts of extreme observed precipitation. This phenomenon, known as "drizzle bias", specifically refers to the model's tendency to overestimate the occurrence of light precipitation events. Consequently, even though the overall precipitation totals are generally well represented, there is often a significant bias in the number of rainy days. The present study aims to minimize the drizzle bias in model output by developing and applying two statistical approaches. In the first approach, the number of rainy days is adjusted based on the assumption that the relationship between observed and simulated rainy days remains the same in time (thresholding). In the second, a machine learning method (random forest or RF) is used for the development of a statistical model that describes the relationship between several climate (modelled) variables and the observed number of wet days. The results demonstrate that employing a multivariate approach yields results that are comparable to the conventional thresholding approach when correcting sub-periods with similar climate characteristics. However, the importance of utilizing RF becomes evident when addressing periods exhibiting extreme events, marked by a significantly distinct frequency of rainy days. These disparities are particularly pronounced when considering higher temporal resolutions. Both methods are illustrated on data from three EURO-CORDEX climate models. The two approaches are trained during a calibration period, and they are applied for the selected evaluation period. [ABSTRACT FROM AUTHOR]

Global and regional climate models' prediction accuracy is limited, often with systematic biases between the model output and observed conditions. The present research aims to minimize the biases between the maximum temperature simulated by a global (HadGEM3) and a regional (WRF) climate model for the period of 2006–2014, based on the ERA5 reanalysis data, used as a proxy for observations (calibration period 1981–2000). For the bias correction, a new approach with the TIN‐Copula method (combination of Triangular Irregular Networks‐TINs and Copulas), suitable for gridded datasets is used. The methodology applies TINs to define the grids to which the corrections are made, and copulas for analysing the dependence between the reanalysis and simulated data during the calibration period. The dependence structure is then used for the bias adjustment. Our methodology is applied to the Middle East and North Africa MENA‐CORDEX domain, where extreme heat conditions prevail, which are projected to accelerate during the 21st century. Besides to the entire area, a detailed analysis was performed for six subregions. Our results indicate that the maximum temperature output of both models diverges from the ERA5 reanalysis, especially during summer. In several subregions and seasons, discrepancies are highest for the HadGEM3 model, which has a much coarser resolution than WRF. The effectiveness of the TIN‐Copula method was tested spatially, temporally and based on two relevant climate indices [daily maximum temperature (TXx) and warm spell duration index (WSDI]), showing that it minimizes model biases. [ABSTRACT FROM AUTHOR]

One of the most common questions in hydrological modeling addresses the issue of input data resolution. Is the spatial analysis of the meteorological/climatological data adequate to ensure the description of simulated phenomena, e.g., the discharges in rainfall–runoff models at the river basin scale, to a sufficient degree? The aim of the proposed research was to answer this specific question by investigating the response of a spatially distributed hydrological model to climatic inputs of various spatial resolution. In particular, ERA-Interim gridded precipitation and temperature datasets of low, medium, and high resolution, i.e., 0.50° × 0.50°, 0.25° × 0.25°, and 0.125° × 0.125°, respectively, were used to feed a distributed hydrological model that was applied to a transboundary river basin in the Balkan Peninsula, while all the other model's parameters were maintained the same at each simulation run. The outputs demonstrate that, for the extent of the specific basin study, the simulated discharges were adequately correlated with the observed ones, with the marginally best results presented in the case of precipitation and temperature of 0.25° × 0.25° spatial analysis. The results of the research indicate that the selection of ERA-Interim data can indeed improve or facilitate the researcher's outputs when dealing with regional hydrologic simulations. [ABSTRACT FROM AUTHOR]