A spatiotemporal analysis of precipitation anomalies using rainfall Gini index between 1980 and 2022

As a reaction to the expanding challenges associated with social susceptibility and their interconnection to diverse environmental threats, parametric insurance plays a key role as an innovation tool in the insurance sector to enhance social resilience to natural disasters and extreme climatic conditions, which can tremendously impact several economic sectors, including agriculture and as a result food security. In this context, this research investigates the association between rainfall Gini index and drought events in Western Europe. For this purpose, we acquired ERA5 data for daily precipitation for five locations from 1980 to 2022. Gini index (GI) values were calculated and analyzed for each location with the Mann–Kendall test at a 5% significance level. As expected, a minimal decreasing trend has been observed for daily precipitation, while an increasing trend was recorded for Gini index. In addition, data on the soil moisture index (SMI) and top drought events were extracted from the European Drought Observatory (EDO) to explore their potential connection with the Gini index over time and space. Although a moderately low to negligible correlation, ranging between −0.27 and 0.02, was found between SMI and GI, a qualitative comparison between major drought episodes and Gini index anomaly showed that similar spatiotemporal patterns are present across the region, particularly for extreme drought events in 1996–1997 and 2003. The current study elucidates the rainfall Gini index's efficiency as a drought indicator for qualitative analysis, yet more work must be conducted to quantitatively evaluate its association with drought magnitude.

can tremendously impact several economic sectors, including agriculture and as a result food security. In this context, this research investigates the association between rainfall Gini index and drought events in Western Europe. For this purpose, we acquired ERA5 data for daily precipitation for five locations from 1980 to 2022. Gini index (GI) values were calculated and analyzed for each location with the Mann-Kendall test at a 5% significance level. As expected, a minimal decreasing trend has been observed for daily precipitation, while an increasing trend was recorded for Gini index. In addition, data on the soil moisture index (SMI) and top drought events were extracted from the European Drought Observatory (EDO) to explore their potential connection with the Gini index over time and space. Although a moderately low to negligible correlation, ranging between À0.27 and 0.02, was found between SMI and GI, a qualitative comparison between major drought episodes and Gini index anomaly showed that similar spatiotemporal patterns are present across the region, particularly for extreme drought events in 1996-1997 and 2003. The current study elucidates the rainfall Gini index's efficiency as a drought indicator for qualitative analysis, yet more work must be conducted to quantitatively evaluate its association with drought magnitude.

K E Y W O R D S
daily precipitation, drought, ERA5, rainfall Gini index, soil moisture index 1 | INTRODUCTION For many hydro-climatological investigations, precipitation is a crucial meteorological metric to determine regional and global climate change risks (Soltani et al., 2016). As a matter of fact, precipitation depicts a primary hydrologic factor in the land surface water balance, where it replenishes soil moisture, supports vegetation growth, and maintains surface flows and groundwater resources, linking atmospheric and land surface processes together (Abed-Elmdoust et al., 2016;Alimonti et al., 2022;Sarker et al., 2019). Therefore, assessing precipitation patterns and distribution is an essential task toward mitigating the impacts of climate change and improving resilience to extreme weather events and natural disasters. By improving our understanding of precipitation patterns and distribution, we can develop better tools and strategies for managing water resources and reducing drought risks.
Precipitation anomalies induced by climate change tremendously influence crop yield and environmental sustainability, which affect food security worldwide (Zaveri et al., 2020). As a result, understanding its significant impacts on agricultural productivity has recently become crucial as climate change accelerates, although rainfall anomalies have proven to be a severe barrier to human civilization for centuries (Coughlan de Perez et al., 2019;Singh et al., 2018). Changes in daily precipitation distribution have already been picked up in observational records, given that global warming reached 1.5 C in 2022 (IPCC, 2022). For instance, an observational study by Rajah et al. (2014) found global trends in extreme and light precipitation events between 1976 and 2000 using data from 12,513 ground-based stations. Similarly, several regional studies have detected changes in precipitation frequency in Europe and worldwide (Masson-Delmotte et al., 2005;Zeder & Fischer, 2020;Huo et al., 2021), which could be associated with climate change.
As a result of severe climate change, extreme drought events are expected to increase in both frequency and magnitude (Seneviratne et al., 2021), potentially necessitating the incorporation of suitable parametric insurance strategies to mitigate its financial impacts on agricultural production, allowing farmers to recover from the disaster at a faster pace (Broberg, 2019). Parametric insurance is a tool for reducing drought impacts in agriculture, serving as a self-sufficient mitigation mechanism to attenuate social vulnerability to environmental disasters from a broad perspective (Radu & Alexandru, 2022). Typically, it is based on robust indices that have been demonstrated to be correlated with natural disasters (Monteleone et al., 2020), such as drought. Incidentally, the Gini index concept for evaluating rainfall inequalities over a defined period has been used in climate change-related studies (Sangüesa et al., 2018;Voskresenskaya et al., 2013); however, no research has been conducted to study its potential association with drought on the regional scale. Therefore, this research examines daily precipitation changes in five locations in Western Europe between 1980 and 2022. Once precipitation trends and the rainfall Gini index were determined, a correlation analysis was conducted to explore the association between soil moisture index (SMI) and precipitation anomalies between 1995 and 2019. In the present study, Europe's top drought events during the past 42 years have been investigated and linked with the rainfall Gini index anomaly to comprehend its efficiency as a drought indicator for qualitative assessment.
2 | DATA AND METHODS 2.1 | Daily precipitation data ERA5 provides hourly estimates of many atmospheric, land, and oceanic climate variables produced by the European Center for Medium-Range Weather Forecasts (ECMWF) in the Copernicus Climate Change Service framework (C3S, 2022. Data encompass the planet on a grid of 30Ákm, using 137 levels to characterize the atmosphere from the ground up to an 80Ákm altitude. Reduced spatial and temporal resolution information for all variables is included in ERA5 data (Hersbach et al., 2019;Hersbach, Bell, Berrisford, Dahlgren, et al., 2020). Monthly updates of ERA5 from 1959 to the present are provided in real-time within 3 months, while the preliminary daily updates are available within 5Ádays (Hersbach, Bell, Berrisford, Hirahara, et al., 2020). More detailed information about the ERA5 data can be found in ECMWF (2022) documentation. This analysis was conducted from 1980 to 2022 for five locations in Eastern France, as illustrated in Figure 1, with an average record length of 42 years. This period was chosen because precipitation is highly variable and depends not only on climate type and geography (Ghorbani et al., 2021) but also on large-scale weather patterns, which require several years to be assessed (Ali et al., 2003). Moreover, it is the most prolonged period of well-represented ERA5 data compared to the period between 1959 and 1979, as stated in the Climate Data Guide (2022).

| Gini index concept
Gini index, which ranges between zero and one, with zero indicating a uniform distribution of precipitation across all days within a year while one indicating precipitation anomalies, offers a measure of how uneven rainfall is distributed over a period of time (Duan et al., 2022). Before that, it was initially developed by the Italian statistician Corrado Gini in 1912 and adopted for economics to assess wealth disparity (Ceriani & Verme, 2012), but more recently, it has been proposed in other scientific fields like hydrology (e.g., Masaki et al., 2014) and geology (e.g., Edgar et al., 2020). As variability parameters, like standard deviation and variance, are sensitive to the scale and probability of data distribution, the Gini index is more robust and nondimensional, making it less complicated to use under diverse scenarios (Monjo & Martín-Vide, 2016).
Once daily precipitation is acquired, cumulatively aggregated, and normalized to a proportion of total precipitation, the Lorenz curve, a graphical representation of inequality degree in data distribution, is used to determine the Gini index value (Gastwirth, 1972). The Gini index is obtained by multiplying the region between the Lorenz curve and the 45 line by two, representing a uniform precipitation distribution, as expressed in Equation (1).
where G is the Gini index, n is the total number of days, and y i is the precipitation for the i th day.
A spatiotemporal analysis of Gini index results showed that the index is susceptible to wet days number in a given year (Sangüesa et al., 2018). Therefore, the relationship between the Gini index and SMI was evaluated using Pearson Correlation via RStudio. Moreover, Gini index distribution was compared with drought events, expecting a connection between drought and high Gini index values. To account for limited data availability, the consistency of the results was assessed using two time periods: the SMI spanning from 1995 to 2019 and the most significant drought events occurring between 1980 and 2016.

| Temporal trend in daily precipitation and Gini index
Before assessing the temporal trend in daily precipitation data, we checked for autocorrelation in time series data, which is crucial as it can affect the accuracy of statistical inference. Autocorrelation occurs when time series F I G U R E 1 Geographic location of the selected test plots.
values at a given point in time are related to their past values. This can result in nonindependence between observations, which can cause problems in estimating statistical models and calculating hypothesis tests. Therefore, we used the autocorrelation function (ACF) within the "stats" package in RStudio. Based on the test results, it can be concluded that there is no significant autocorrelation in the data since the autocorrelation coefficient ranged on average between À0.005 and À0.02 at a 5% significance level. Then, the Mann-Kendall statistical test was used at all five locations to evaluate trends in daily precipitation and the Gini index at a 5% significance level. The nonparametric test does not presume anything about the data's marginal probability distribution, ascertaining data independency to time (Sangüesa et al., 2018;Song et al., 2015). The Mann-Kendall Trend test's fundamental principle is to compare the signs of earlier and later data points. Therefore, if a trend is present, the sign values will continuously increase or decrease (Meals et al., 2011). Detailed information about the test can be found in Karmeshu (2012) and Wang et al. (2020). Khambhammettu (2005) claims that the data values are assessed as an ordered time series. All following data values are compared to each value. The Mann-Kendall statistic, S, is initially assumed to have a value of zero, then increased by one if a data value from a later period is more significant than a data value from an earlier period, and vice versa. The final value of S is the sum of all increments and decrements.
Given data points x 1 , x 2 , … x n , x j represents the data point at time j. The Mann-Kendall statistic (S) is determined by Equation (2). where: A positive value of S suggests an upward trend, whereas a negative value indicates a downward trend. The probability associated with S and the sample size n must be determined to assess the trend significance statistically.
Then, the variance VAR(S) is determined, as expressed in Equation (3).
Finally, the Z value is calculated depending on S and its variance.

| Soil moisture index
Along with precipitation and evapotranspiration, soil moisture is a key factor in plant growth and a fundamental part of the hydrological cycle. The SMI is defined as the ratio of the difference between present soil moisture and the permanent wilting point to field capacity and residual soil moisture (SMI). A score of zero indicates arid conditions, whereas a score of one indicates extremely wet conditions (Saha et al., 2018;Zhang et al., 2017). The Copernicus European Drought Observatory (EDO) uses SMI to track the beginning and end of agricultural drought conditions, which appear when soil moisture availability to plants falls; hence adversely affects crop output and, consequently, agricultural productivity (Cammalleri et al., 2016). The SMI retrieved from EDO is produced by the joint research center (JRC) using LISFLOOD hydrological model with a spatial resolution of 5Ákm (Van Der Knijff et al., 2010) relative to a climatological reference period (1995-2022) (De Roo et al., 2000). The drought indicator assesses one of the three primary droughts-agricultural drought-based on the hydrological cycle variables (Laguardia & Niemeyer, 2008). Agricultural drought results in decreased crop production due to insufficient soil moisture, while hydrological drought is characterized by below-average water availability in rivers, streams, reservoirs, lakes, or groundwater, and both droughts are typically preceded by a meteorological drought, which is a prolonged period of less than average rainfall in a given region. More details about SMI can be found in the EDO indicator factsheet (EDO, 2021). SMI is determined using Equation (5), with θ being the daily soil moisture and θ 50 being the average between the wilting point and the field capacity.
We did not normalize SMI as suggested by a few studies (e.g., Grillakis (2019)) to maintain its seasonal representation, as SMI is merely a measure of soil moisture, reflecting the seasonal patterns of precipitation and evapotranspiration. Therefore, removing the inherent seasonality would fundamentally change the significance of the SMI and may not accurately reflect the underlying dynamics of the soil moisture system, in other words, to preserve the temporal patterns since seasonality is a critical factor in the water cycle and influences the temporal variability of moisture. Modifying the SMI by removing the latter may mask significant temporal signals related to precipitation and evapotranspiration, which are crucial to characterizing the water cycle and monitoring drought conditions.

| Top drought events database
The Global Drought Observatory provides free access to its drought events database, which was compiled by Spinoni et al. (2019) using over 4500 global drought events that occurred between 1951 and 2016. At various temporal scales, they have been categorized using a few metrics, including the standardized precipitation index (SPI) (McKee et al., 1993) and the standardized precipitation evapotranspiration index (SPEI) (Vicente-Serrano et al., 2010), which are the most commonly used multiscalar indicators based on climatic data to determine meteorological droughts. Even though the standardized precipitation evapotranspiration index (SPEI) is a refined version of the widely used standardized precipitation index (SPI), it has the advantage of considering not only precipitation but potential evapotranspiration (PET) when evaluating drought. A detailed description of SPI and SPEI can be found in EDO's indicator factsheet (2021).
The database contains separate entries for around 175 countries and macroregions on a spatial scale; among them, 52 of the largest mega-droughts occurred between 1951 and 2016 using a specific classification scoring system (Spinoni et al., 2019). Moreover, according to the database, Amazonia, southern South America, the Mediterranean region, much of Africa, and north-eastern China stand out as drought hotspots. With rising temperatures over North America, Central Europe, Central Asia, and Australia, precipitation anomalies increased, leading to frequently severe droughts, which will be investigated in this study.
The drought events database is updated every 2Áyears and serves as a valuable resource for both scientific investigations and policy reports. This study has incorporated the database to compare drought patterns between 1980 and 2016 with the spatiotemporal distribution of the Gini index in the selected region. Further, this study aims to qualitatively examine the interconnection between yearly precipitation inequality and the probability of a drought event occurring in the same year.
3 | RESULTS AND DISCUSSION 3.1 | Descriptive analysis Table 1 illustrates the main statistical parameters for daily precipitation in selected locations over the last 42Áyears. From 1980 to 2022, significant increases and decreases in rainfall distribution were recorded. The distribution of daily precipitation values is characterized by a significant difference between the minimum and maximum values, reflecting this parameter's spatial variability. Further, a discard between the mean and median values has been observed for all locations, revealing a positive skewness in time-series data. Minimal decreasing trends were predominantly observed with a tau value ranging between À0.259 and À0.256 at a 5% significance level. This decreasing trend is mainly consistent with the European Environment Agency report published in 2017 (EEA, 2017), which revealed a decreasing trend in annual precipitation across western and southern Europe between 1960 and 2015.
The main statistical parameters for the Gini index are given in Table 2. A slight difference between the minimum and maximum values was observed, indicating a spatial variability of the Gini index over time. Plus, the normality test demonstrated that the index is normally distributed with a remarkably close range between mean and median values. The coefficient of variation (CV), which is a relative measure of data distribution expressed as the fraction of standard deviation to mean, ranged between 3.36% and 3.69% for all test locations, revealing the consistency and reliability of measured values. The CV can help compare variation degrees from one dataset to another, which was insignificant in this case. Based on Table 2 and Figure 2, increasing trends were observed in Gini index variation over time for test locations, with a tau value ranging between 0.18 and 0.3 at a 5% significance level. The positive trend is somehow in agreement with Rajah et al. (2014), who discovered a significantly decreasing frequency of wet days in the same region between 1976 and 2000. Given that the Gini index is a direct indicator of rainfall distribution inequality over time, this change in frequency has led to a positive trend.

| SMI and GI
Using RStudio, a Pearson correlation between SMI anomaly values and the Gini index for each station was separately conducted to evaluate precipitation anomalies' effect on soil moisture. Correlation between the GI and SMI anomaly retrieved from the EDO portal shows a negligible to moderately low correlation varying between À0.27 and 0.02, as illustrated in Figure 3. To some extent, the correlation was expected to be moderately negative, which occurred only in two locations (i.e., Monestier-Port-Dieu and Les Challots). Since soil moisture patterns depend not only on precipitation variability but also on evapotranspiration and runoff (McCabe & Wolock, 2013), connections between soil moisture as a feedback mechanism for precipitation and other processes are not simply linear; hence soil moisture dependency on precipitation among other hydrologic components is to be studied (Liang et al., 2010). Further, soil moisture is affected by other pedogenic factors like soil texture, composition, and organic matter content, as well as geomorphic and topographic properties like land cover and altitude, which agrees with Sehler et al. (2019), who investigated the complex relationship between soil moisture and precipitation using remote sensing data. The study found that soil moisture and precipitation have a higher correlation under sparse vegetated dry areas, whereas forests and densely vegetated regions have a weaker relationship, which is the case of selected locations, as illustrated in Figure 1.
Although the correlation between SMI anomaly and Gini index is extremely low, a qualitative analysis of graphs presented in Figure 4 revealed a sort of synchrony, to some extent, between Gini index and SMI anomalies. Over time, variations in SMI anomaly coincide with a rise in Gini index anomaly resulting in a local maximum. For instance, a peak in GI anomaly at Monestier-Port-Dieu in 1998, 2000, 2004, and 2013 is synchronized or shortly followed by a gradual change in SMI anomaly. This is even more prominent in these stations, that is, Bief de Pont-Loup, Les Challots, and Vallée des Gobillons, for the same period. The contemporaneous nature of this phenomenon may be attributed to the complex interactions between the various components of the hydrological cycle, spanning from the atmosphere to the soil, across both space and time. These findings align with previous research by Meng et al. (2018) and Wang et al. (2018). However, it is essential to note that these interpretations are solely based on present observations, and further research can be carried out to explore this aspect in greater detail.
While SMI can be a reliable index for assessing drought events, it has a few limitations to be considered when interpreting its values and integrating it into decision-making regarding resource management and planning. SMI only provides information about the moisture content of the soil's top layer and does not consider the deeper layers, which can also be essential for agricultural and ecosystem processes (Dai & Trenberth, 2004). In addition, it is often calculated using remote sensing data, which can have limitations in terms of spatiotemporal resolution (Schwank et al., 2013). For instance, remote sensing data may not accurately reflect the spatial variability of soil moisture within a given area, as it can be easily affected by

| Top drought events and Gini index
As a severe natural hazard, drought has occurred in diverse macroregions from arid to subhumid climates (Carrão et al., 2016). This has attracted researchers' attention to study its spatial distribution, origins, and impacts in order to mitigate its risks (Mishra & Singh, 2010). Many factors play a significant role in drought, like extreme heat waves, low relative humidity, and rainfall frequency and distribution during the growing season (IDMP, 2022). However, low-rainfall areas are the most vulnerable spots to drought episodes primarily due to F I G U R E 2 The temporal trend of the rainfall Gini index in the test locations based on the Kendall-Mann test over 42Áyears.
In this context, the top drought events in Europe between 1980 and 2015 were extracted and compared with the Gini index variation, as presented in Figure 5. Drought episodes have consistently occurred alongside local peaks in the Gini index during the same year, indicating a potential association between drought and precipitation, which is in line with the findings of Wang et al. (2021) and Moazzam et al. (2022) regarding the effect of precipitation heterogeneity on the occurrence of seasonal droughts. The non-synchrony cases between drought and Gini index peaks can be explained by the fact that precipitation anomalies were not the main driver for the drought episode like the case of 1991-1992, which was a prolonged drought period across the F I G U R E 3 Correlation between Gini index and soil moisture index anomaly in different test areas.
Mediterranean due to rain deficit in that region according to the European Drought Center report (1995). Nevertheless, it probably extended to surrounding areas, mainly Northern Europe.
Furthermore, the spatiotemporal variation of drought events has been investigated. As shown in Figure 6, the most significant drought episodes occurred in 1982, 1985, 1992, 1993, 1996-1997, 2003, 2011, and 2015, with 1996-1997 and 2003 being the most severe, followed by 2011. According to the European Drought Center, the 1996-1997 drought was characterized by meteorological shortages in the UK, followed by an ever-expanding drought in Southern Scandinavia and Northern Germany (EDC, 1997). The event began with a dry winter in 1995 and continued into 1996 and the spring of 1997, reducing available water for western Europe in the early spring. This is consistent with the Gini index anomaly map for the same period, as shown in Figure 7.
In addition, a large part of Europe was affected by another severe drought in 2003 as a consequence of rain deficits combined with elevated temperatures, causing high evapotranspiration levels that led to around ten billion Euros of agricultural losses, according to the EDO Drought News report released by the joint research center (2015). This is consistent with our results of the Gini index anomaly for the same year, which reached +0.2 across Western Europe. From January to April 2011, severe cumulated rain deficits were recorded in France and the rest of Europe, considered the driest period since 1975, according to the EDO Drought News report of 2011 (EDO-JRC, 2011). In fact, similar patterns were found in the 2011-Gini index map. Besides, the prolonged rainfall shortages in April 2015 and the temperature anomalies in July of the same period caused a severe drought that affected soil moisture content and vegetation (EDO-JRC, 2015), in consistency with the Gini index map for 2015.
For validation purposes, we have compared the patterns of Gini index anomaly in 2021 and 2022 with the occurrence of drought events in the same period. Although the 2021 spring in Europe was cooler than usual, with certain regions experiencing an early spring followed by a late frost episode that impacted agricultural productivity hence crop yields, the summer brought record-high temperatures, intense and protracted heatwaves, and flooding events, making it the "year of contrasts" for Europe between heatwaves, wildfires, and floods. The European State of the Climate (2021) report states that outside of the 10 warmest years on record, annual surface air temperatures in 2021 were only about 0.2 C higher than the average from 1991 to 2020. Indeed, this corroborates with the regional pattern found in the Gini index anomaly map of 2021, which revealed an abundance of negative anomalies that could reach À0.05 in some areas, as shown in Figure 8. In contrast to 2021, according to Toreti et al. (2022), a severe drought affected Europe since the beginning of 2022 and continues to expand in July and August 2022. The main drivers are lack of rainfall which mainly affected water resources by reducing river discharges in Europe, and combined with intense heatwaves, which led to a deficit in crop yields in May and June 2022. As a result of water resources shortages, severe impacts on the energy sector have been registered, as well as significant losses in agricultural productivity of maize, soybeans, and sunflowers fields. The overall situation agrees with the spatial pattern in the Gini index anomaly map for the same year, as presented in Figure 9.
Merely, the drought scenarios explained above compared with Gini index results prove the potential of the Rainfall Gini index as a valuable drought indicator that can be considered in future studies related to drought monitoring and mapping. This can be further confirmed by Samantaray et al. (2022) regarding rainfall patterns' connection to drought characteristics. Further, Nugroho et al. (2021) revealed that precipitation anomaly is a triggering factor for drought based on a statistical analysis conducted across Indonesia using Climate Hazards Group InfraRed Precipitation with Station (CHIRPS) F I G U R E 5 The variation of the rainfall Gini index in the test locations in synchrony with the major drought events recorded since 1980, according to the European drought observatory (EDO), where the vertical lines symbolize drought episodes. data, which corroborates with the present study regarding the present synchrony between precipitation anomaly and drought occurrence over practically the same period.
While our comparison was qualitative and did not assess the severity or magnitude of temporal changes, the synchronicity between the Gini index and drought is visually evident and firmly established. However, a statistical analysis of the relationship between these two factors would be necessary to draw conclusive evidence. This is consistent with the study of Luo et al. (2017), who found remarkably comparable results between precipitation anomalies and long-term agricultural drought episodes in California from 2012 to 2015. In a similar context, De la Barreda et al. (2020) also investigated drought episodes in the Yucatan Peninsula (Mexico) in relationship with precipitation anomalies between 1980 and 2011. The study found that drought can be regionalized using information about precipitation anomalies over time to improve drought monitoring and land management strategies in the area, which agrees with the current study's aims and findings.
It is worth noting that the relationship is not very straightforward to be outlined due to the complicated dynamics of drought and its association with multiple factors, not only rainfall but other diverse biotic and abiotic factors like temperature, soil, and vegetation. Nevertheless, this qualitative analysis reveals that the Gini index can be a valuable tool for drought monitoring and assessment while considering other influential factors. As this indicator had comparable regional and temporal patterns, it could be linked to a simultaneous occurrence of a drought season in the exact location. Many scenarios could be investigated using the Gini index as it clearly showed related patterns to drought. Future work will explore the statistical aspect of this connection and quantify drought risks and magnitude using Gini index anomaly and other related parameters like soil type, vegetation, land use and land cover, and topography.

| CONCLUSION
The results indicated that daily precipitation oscillated over time with a minimal decreasing trend from 1980 until 2022, revealed by a tau value ranging between À0.259 and À0.256 at a 5% significance level. Further, the Gini index showed an increasing trend (0.18-0.3), with an overall low negative correlation with the SMI (4 out of 5 selected locations). This suggests that each selected location had particular environmental characteristics and local climatic conditions, making it more challenging to generalize. Equally, soil moisture heavily depends on various environmental, pedogenetic, topographic, and geomorphologic features, not solely precipitation, making it more challenging to simplify interactions among various hydroclimatic components to a single variable. On the other hand, not only did the Gini index often show synchrony with drought events over time, but its spatial variation revealed similar patterns with significant drought events in Western Europe.
Although there was no statistical significance for the relationship between Gini index and SMI, comparing patterns of drought extent with Gini index showed that the Gini index can be considered a reliable drought indicator in qualitative analysis; hence can be used in parametric insurance for drought events assessment and monitoring. Future studies should focus on quantitatively establishing the statistical relationship while considering other key drivers.
This study proves that the rainfall Gini index can be used to understand drought events in Europe. The uniqueness of the research is reflected in applying it as an indicator of drought events instead of traditional indicators like the SMI, emphasizing its potential in drought monitoring and early warning systems. The increasing trend in the Gini index can be used to monitor and predict the likelihood of drought events, which can impact agriculture and food security on regional and global scales.
F I G U R E 9 Gini index anomaly map for 2022. resources; software; validation; writingreview and editing.