Cross-comparison of national drought monitoring products in Central Europe using a new drought impact database

Abstract Monitoring drought helps to reduce their economic and environmental impacts by enabling early warnings and better resource management planning. In Europe, there are several operational monitoring systems operating at national and regional scales. However, such monitoring systems are rarely validated, which complicates the decision-making process. Therefore, we evaluated six national drought monitoring products in Central Europe using a novel extreme event impact database compiled from national newspaper reports over the period 2000–2023. The drought monitoring indices used in the countries include the standardized precipitation index (SPI), standardized precipitation evapotranspiration index (SPEI), and standardized relative soil moisture with different aggregation periods. The area under receiver-operating characteristic curve (AUC) is used to assess the ability of the drought indices to detect impact occurrence. Spearman correlation coefficients (r) between the severity of the drought index and the number of reported impacts are used to assess their ability to capture impact severity. The highest AUC values were obtained for the drought monitoring products of Czechia, Croatia, and Slovenia ( $$AUC&gt;0.8$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>A</mml:mi> <mml:mi>U</mml:mi> <mml:mi>C</mml:mi> <mml:mo>&gt;</mml:mo> <mml:mn>0.8</mml:mn> </mml:mrow> </mml:math> ) while the lowest values were obtained for the monitoring product of Austria ( $$AUC&lt;0.7$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>A</mml:mi> <mml:mi>U</mml:mi> <mml:mi>C</mml:mi> <mml:mo>&lt;</mml:mo> <mml:mn>0.7</mml:mn> </mml:mrow> </mml:math> ). Impact severity was best captured in Poland (for some indices $$r&gt;0.6$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>r</mml:mi> <mml:mo>&gt;</mml:mo> <mml:mn>0.6</mml:mn> </mml:mrow> </mml:math> ), and worst in Slovakia, Slovenia, and Austria ( $$r&lt;0.4$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>r</mml:mi> <mml:mo>&lt;</mml:mo> <mml:mn>0.4</mml:mn> </mml:mrow> </mml:math> ). With an increasing aggregation period, the correlation generally decreases, while the AUC values show a non-linear pattern, peaking at an intermediate integration time of three to 6 months. The results of this study help to understand the strengths and weaknesses of drought monitoring products in each country and support the development of a common drought monitoring framework for Central Europe.