Characteristics of Meteorological Droughts Across Different Climatic Zones in Benin

Abstract

This study investigates meteorological drought characteristics across three climatic zones in Benin using the SPEI (Standardized Precipitation Evapotranspiration Index) and SPI (Standardized Precipitation Index). A comprehensive statistical approach was employed, including the Mann–Kendall trend test, drought duration and intensity analysis, Pearson correlation, cross-wavelet transform, and the Standardized Relative Air Humidity Index (SRHI), to assess drought patterns and trends. The findings indicate increasing consistency between SPI and SPEI trends at longer timescales, though significant regional variations persist. In Zone 1 (northern Benin), SPI exhibited an increasing trend across all timescales, whereas SPEI demonstrated a decreasing trend at shorter timescales. In contrast, in Zones 2 (central Benin) and 3 (south Benin), both indices generally displayed a decreasing trend, except at the one-month scale. An analysis of drought duration and intensity revealed that, at shorter timescales (SPI and SPEI at 1- and 3-month intervals), the longest droughts occurred in Zones 1 and 3, while the most intense events were recorded in Zone 2. At longer timescales (SPI and SPEI at 6- and 12-month intervals), Zone 2 experienced the longest droughts, whereas Zone 3 exhibited the highest intensities. These findings illustrate the need for monitoring strategies tailored to a given area’s characteristics. Despite these insights, data uncertainties and regional differences present challenges for drought investigation. Future studies should incorporate more datasets and investigate different drought indices to improve decision-making and improve strategies for safeguarding Benin’s agricultural sector, ecosystems, and food supply.

1. Introduction

Drought is one of the worst natural calamities, which presents serious problems for agriculture, the economy, and society [1]. While there is no universally accepted definition of drought, it is often characterized by an abnormal increase in temperature or a reduction in precipitation levels relative to historical norms [2,3]. Drought can also be characterized as a lack of soil moisture or a sustained scarcity of water due to months, seasons, or even years of inadequate rainfall [4,5]. Because drought events grow gradually and are complicated, it can be challenging to anticipate when they will start and stop [5,6].

Drought is generally categorized into four distinct types: (i) Meteorological drought, which is characterized by a prolonged lack in precipitation and an extended dry period; (ii) Hydrological drought, which arises from persistent precipitation shortages, leading to decreased water availability in rivers, lakes, and other water bodies; (iii) Agricultural drought, which takes place when a lack of rainfall, depletion of soil moisture, or declining groundwater and reservoir levels negatively affect crop growth and livestock farming [1,7]; and (iv) Socio-economic drought, which arises when water shortages disrupt the supply of essential goods such as agricultural produce and livestock products, thereby affecting economic stability and livelihoods https://www.weather.gov/safety/drought (accessed on 15 November 2024). West Africa is one of the most climate-sensitive regions in Africa, with vulnerability exacerbated by multiple stressors at various scales and limited adaptive capacity among local populations [7]. Approximately 33% of Africa’s population lives in regions vulnerable to drought, making them highly susceptible to its adverse effects [8]. In Benin and its neighboring countries, recurrent droughts and floods significantly impact communities, livelihoods, and ecosystems [9]. Given these challenges, a comprehensive comparative analysis of appropriate drought assessment tools is essential for improving monitoring, mitigation, and adaptation strategies in Benin.

Researchers frequently utilize drought indices to evaluate and describe drought events, as these indices offer crucial information on climatic variables such as intensity, severity, and duration within a specific region [10,11]. These indices, grounded in the water balance theory, incorporate key hydrometeorological components such as temperature, runoff, soil moisture, and rainfall to enhance drought monitoring and analysis [12]. Commonly used indices include the SPI [13], the SPEI [14], SRHI [15], along with the Reconnaissance Drought Index (RDI) [1] and the Composite Meteorological Drought Index (CI) [16]. Additionally, indices such as the Soil Moisture Drought Index (SMDI) [17], the Standardized Runoff Index (SRI) [18], and the Palmer Drought Severity Index (PDSI) [19] incorporate hydrological and climatic variables, including precipitation, soil moisture, runoff, and potential evapotranspiration, to assess drought conditions [12]. These indices play a crucial role in evaluating drought characteristics and informing water resource management and climate adaptation strategies.

This study selected the SPI, SPEI, and SRHI, considering their respective characteristics, relevance to local environmental conditions, and the availability and reliability of data. The SPI is the most widely adopted meteorological drought index due to its simplicity, efficiency, and ease of calculation, as it relies on long-term precipitation records (typically over 20 years) [1,19]. SPI calculations typically assume that rainfall and other meteorological variables remain constant over time, without exhibiting any temporal trends. In contrast, the SPEI, as outlined by Vicente-Serrano et al. [14], builds upon the SPI by integrating evapotranspiration, allowing for a more comprehensive drought assessment that accounts for both precipitation and temperature influences. SPEI captures the multi-temporal nature of drought events, calculated at various timescales, depending on precipitation probability and differences in potential evapotranspiration. Both SPI and SPEI can be computed for different timescales—commonly, 1, 3, 6, 12, and 24 months [20,21]. Short-term scales, such as 1, 3, and 6 months, are particularly relevant for monitoring agricultural drought impacts, while longer scales of 12 and 24 months are crucial for understanding hydrological and socio-economic drought impacts. This flexibility makes SPI and SPEI well-suited for comprehensive drought analysis in Benin’s diverse climatic zones. RH is one of the climate variables that is easier to measure and, hence, RH observation data are readily available in Benin. RH was used to identify which of the two indices is appropriate for the monitoring of agricultural drought in the different zones of Benin. Various studies have shown that reliable results can be obtained from the application of SRHI for monitoring agricultural drought; this is the case in [10,22,23].

The Mann–Kendall test, a commonly applied non-parametric statistical approach, has been applied in numerous studies across different regions to evaluate the performance of various drought indices [11,16]. For instance, ref. [14] analyzed time-series data of the SPI, the SPEI, and the PDSI across multiple observatories with diverse climatic conditions. Their findings suggested that SPEI and PDSI were more effective for drought monitoring and analysis. Similarly, ref. [24] assessed the performance of SPI and SPEI over different timescales to determine the most suitable drought index for the Punpun River Basin. In Bangladesh, [25] conducted a comprehensive evaluation of SPI and SPEI and found that SPEI demonstrated superior performance. In the West African region, the SPEI and the SRI were employed to investigate drought characteristics in the Niger River Basin and assess their influence on river flow [26]. Additionally, ref. [27] evaluated SPEI’s effectiveness in capturing historical drought events within specific Nigerian agroecological zones.

More recently, refs. [18,28] analyzed drought characteristics from 1951 to 2100 using multiple indices (SPEI, SRI, RDI, and SPI), emphasizing the necessity of selecting drought indicators that align with regional climate and hydrological conditions. This approach is supported by contemporary studies in West Africa, where comparative analyses of these indices reveal critical insights. For instance, refs. [29,30] demonstrated that SPEI and SRI better capture compound drought risks in the Sahel due to their integration of temperature and runoff dynamics, while ref. [31] found RDI more effective than SPI for agricultural drought monitoring in Sudanian zones. Similarly, ref. [30] highlighted SPEI’s superiority in reflecting warming-induced drought severity, and ref. [32] identified lags between SPI/SPEI (meteorological) and SRI (hydrological) droughts. Together, these studies underscore the importance of index selection tailored to regional drivers, whether precipitation deficits (SPI), evaporative demand (SPEI/RDI), or hydrological responses (SRI), to improve drought early warning and adaptation strategies.

The characterization of droughts and their spatiotemporal variability has been a primary focus of numerous studies in the West African region. For instance, ref. [33] analyzed climate trends and regime shifts within the Sota watershed in Benin using the SPI. Similarly, refs. [23,34,35,36] investigated the temporal and spatial variations in precipitation across eight precipitation stations within the lower valley of the ORB from 1970 and 2016. Furthermore, ref. [37] reported a significant decline in annual precipitation, averaging −5.5 mm/year, along with a notable increase in mean temperature, exceeding 1 °C, across three distinct climatic zones in Benin between 1960 and 2008. These findings underscore the evolving climatic patterns in the region and their potential implications for hydrological and agricultural systems.

It is important to highlight that the aforementioned studies on drought assessment in Benin did not involve a comparative analysis of different drought indices. To establish robust guidelines for choosing the best drought index for Benin, a systematic comparison between the SPI and the SPEI is essential. The main goal of this study is to analyze meteorological drought characteristics across three distinct climatic zones in Benin using the SPI and the SPEI. To achieve this, we employed a range of analytical methods, including temporal variation analysis, Pearson correlation, the Mann–Kendall trend test, drought duration and intensity assessment, and cross-wavelet transformations. Furthermore, we investigated the relationship between the SRHI and both SPI and SPEI to determine the most suitable index for effective drought monitoring in the region.

2. Research Data and Methods

2.1. Study Area

The study focuses on Benin, a West African nation spanning 114,763 km² between longitudes 1° and 3°40′ E and latitudes 6°30′ and 12°30′ N [9]. The elevation across the country ranges from −14 to 671 m above sea level. Benin is traversed by several rivers, as illustrated in Figure 1. The most prominent is the Ouémé River, which is the longest river in the country. It spans approximately 510 km and drains a watershed of about 50,000 km², originating from the Tanéka Mountains in the north and flowing southward to the eastern shores of Lake Nokoué. Benin is home to three major cities: Cotonou, the economic capital and the most populous urban center; Porto-Novo, the official administrative capital, known for its rich cultural heritage; and Parakou, the principal city in the northern region and a vital commercial hub. The country shares its borders with Niger to the north, Burkina Faso to the northwest, Nigeria to the east, Togo to the west, and the Atlantic Ocean to the south, as depicted in Figure 1. This strategic location creates diverse climatic conditions, facilitating comprehensive drought analysis. Benin’s tropical climate system comprises three distinct zones [38]. Zone 1, known as the Sudanian Zone, is located in northern Benin and includes 10 weather stations. This zone experiences a distinct extended dry season (October to May) and a brief rainy season (June to September), with soils predominantly hydromorphic and well-drained, consisting largely of lithosols. The vegetation is primarily composed of savannas with scattered small trees [37]. Zone 2, referred to as the Sudano-Guinean Zone, encompasses central Benin and includes 17 weather stations. This zone has two rainy seasons: rainy May–mid-June and unstable rainfall pattern in September–October. Ferruginous soils of varying fertility sustain mixed vegetation, including open forests and gallery woodlands [37]. Lastly, Zone 3, referred to as the Guinean Zone, located in southern Benin, includes eight weather stations. It is characterized by an equatorial climate with dual rainy seasons (April–July and September to November). This area has soil that is either deep and ferralitic, or rich in vertisols, humus, and minerals, which support diverse vegetations that are adapted to the humid conditions [9]. Meteorological data (precipitation, temperature, relative air humidity) were acquired from Benin’s National Meteorological Agency (NMAB), with 35 stations providing comprehensive coverage across these ecoclimatic regions. This spatial distribution enables a robust evaluation of drought patterns throughout Benin’s heterogeneous landscapes.

Land-Use/Land-Cover

The land-use and land-cover (LULC) maps provide essential information on the physical characteristics and classification of the landscape across Benin. Land is a fundamental natural resource underpinning all human activities, and a comprehensive understanding of land-use types and their spatial distribution is critical for effective spatial planning, sustainable land management, and resource optimization [39]. The LULC data were instrumental in identifying and categorizing major land-cover types throughout the country. Five primary land-use/-cover classes were identified, as illustrated in Figure 2: grasslands (1857.68 km²), shrub-covered areas (50,676.82 km²), tree-covered areas (19,102.9 km²), croplands (43,508.66 km²), and water bodies (616.34 km²). The LULC dataset used in this study was obtained from the FAO data platform: https://data.apps.fao.org/map/catalog/srv/eng/catalog.search#/home (accessed on 15 November 2024).

2.2. Spatial Patterns and Temporal Trends in Precipitation, Temperature, and Relative Air Humidity (RH)

The spatial distribution of average annual temperature, precipitation, and RH across the research area is depicted in Figure 3a–c. Zone 3 receives the least precipitation (1232.74 mm), followed by Zone 2 (1260.14 mm), while Zone 1 records the highest precipitation (1373.8 mm). The annual mean temperatures for Zones 1, 2, and 3 are 28.03 °C, 27.85 °C, and 27.89 °C, respectively. Furthermore, Zone 2 has the highest mean RH (73.55%), while Zone 3 has the lowest (66.05%) in comparison to Zone 1 (68.23%).

Temporal trends indicate a slight decrease in annual precipitation (−0.66 mm/year), as shown in Figure 3d, whereas Figure 3e illustrates a minor upward trend in temperature (0.02 °C/year). Additionally, Figure 3f highlights a marginal decline in RH (−2.94 × 10⁻⁴)%/year). In this study, the weather stations selected for analysis had complete datasets. However, observational data were not available for every site in Benin. To address this spatial data gap, the Inverse Distance Weighting (IDW) interpolation method was applied to estimate unknown values at unsampled locations based on values from nearby known stations. Precipitation values for each zone were obtained by aggregating the precipitation data from individual stations within that zone, while the temperature and RH for each zone were calculated as the average values across the corresponding stations. Figure 4 illustrates the methodological workflow for analyzing meteorological observations and trends in Benin from 1988 to 2021, providing a systematic visualization of climatic data distribution patterns.

2.3. Methodology

2.3.1. Data Source and Processing

The research employed meteorological data obtained from Benin’s NMAB, comprising daily precipitation, temperature, and RH measurements recorded at 35 monitoring stations. While the complete dataset spans 35 years (1987–2021), the analytical timeframe focused on the 34-year period from 1988–2021. Initial 1987 data served as baseline for computing 12-month scale coefficients applicable to January 1988. The dataset includes comprehensive geospatial metadata for all stations, along with daily averaged climatic parameters.

2.3.2. Calculation of the SPI and SPEI

This investigation employed the SPI and SPEI, two globally recognized drought indices recommended by the World Meteorological Organization for monitoring drought conditions worldwide [18,40]. The study’s application of these indices across Benin’s diverse climatic regions provides valuable insights into their performance under varying ecological conditions.

One of the indicators used to show how much precipitation fell globally over a certain time period is the SPI. The two-parameter gamma distribution with a shape and scale parameter is the most widely utilized distribution for SPI computation. The gamma distribution’s probability density function is defined as [3,14]

$$G(x)={\displaystyle \frac{1}{{\beta }^{\theta }\tau (\theta )}}{\int }_0^x{x}^{\theta -1}{e}^{-x/\beta }\mathrm{d}x, \mathrm{for}x>0$$

(1)

where $x$ represents the precipitation value; $\beta$ the scale parameter; $\theta$ the shape parameter; and $\tau \left(\theta \right)$ the gamma function. Precipitation can have values of zero; however the distribution of gamma is undefinable for $x=0$. Therefore, the distribution of cumulative probability for a zero-precipitation value is derived as follows:

$$H\left(x\right)=w+(1-w)G\left(x\right)$$

(2)

where $w$ represents the probability of the zero-precipitation value.

The SPEI considers both precipitation (P) and potential evapotranspiration (PET), while the SPI relies exclusively on precipitation data. PET values are estimated using the Hargreaves method [18,41,42]. The monthly water balance for a given month n is determined by computing the difference between precipitation and PET [18]:

$$D_n=P_n-{PET}_n$$

(3)

where $D_n$ represents the $n_{th}$ month’s moisture deficit (mm). $P_n$ is the $n_{th}$ month’s precipitation (mm) and ${PET}_n$ is the $n_{th}$ month’s potential evapotranspiration (mm). The PET was calculated automatically using the SPEI package (version 1.7) in R 4.2.2 software software (http://cran.r-project.org/web/packages/SPEI, accessed on 15 November 2024) based on the Hargreaves method.

To compute the SPEI, $D_n$ is fitted to a three-parameter probability distribution [11,43]. The calculated SPEI values across different timescales are aggregated using the same methodology as the SPI. This study utilized the ClimPACT2 package in the R programming environment, available at (https://github.com/ARCCSS-extremes/climpact2/, accessed on 15 November 2024) to derive both SPI and SPEI. The analysis incorporated daily precipitation data along with minimum and maximum temperature records collected over a 36-year period, using 1987–2021 as the reference years.

SPI and SPEI are essential tools for assessing moisture variability and identifying both dry and wet periods. A value below −1 for either index indicates drought conditions, while values above +1 signify wet periods, with more extreme values reflecting greater severity [23]. These indices can be applied across multiple timescales, including 1, 3, 6, 12, and 24 months, allowing for the characterization of different types of drought. Shorter timescales (1–3 months) are commonly used for meteorological drought assessment, whereas agricultural drought is typically analyzed over 3–6 months. In contrast, hydrological drought and water resource availability are better captured over extended periods of 12–24 months [43].

Drought conditions, as determined by the SPI, SPEI, and SRHI indices, are categorized as follows: normal (values ranging from 0 to 0.99), mild drought (values between 0 to −0.99), moderate drought (values from −1.00 to −1.49), severe drought (values between −1.50 to −1.99), and extreme drought (values of −2.00 or lower) [44].

2.3.3. Mann–Kendall (MK) Trend Test

The MK test is a widely used nonparametric method for detecting persistent patterns in climatic and hydrological time series, including runoff, temperature, and precipitation [34,45]. This test is particularly effective for analyzing datasets with non-normal distributions, as it is less sensitive to extreme values and data skewness. The trend’s significance is assessed based on established confidence levels: the trend is considered significant at the 99% confidence level if ∣MK∣ > 2.58|MK| > 2.58∣MK∣ > 2.58 (p = 0.01), at the 95% confidence level if ∣MK∣ > 1.96|MK| > 1.96∣MK∣ > 1.96 (p = 0.05), and at the 90% confidence level if ∣MK∣ > 1.65|MK| > 1.65∣MK∣ > 1.65 (p = 0.1) [46]. The significance of monotonic growing or declining trends in hydro-meteorological series data was assessed in this study using the Mann–Kendall test, as explained by [43]. This comprises several drought indicators for various areas from 1988 to 2021 at different periods (i.e., 1, 3, 6, and 12 months).

2.3.4. Duration and Intensity of Drought Events

Drought indices are essential tools to assess drought impacts and characterize its key attributes, including duration and intensity [9]. This study used SPI and SPEI as the main drought indicators. These indices were computed across various timeframes, including short-term and long-term durations, to provide a comprehensive evaluation of drought characteristics.

Drought duration is characterized by consecutive months where drought index values remain below −1, bordered by periods with values exceeding −1 [7,47]. Drought intensity is measured as the cumulative SPI or SPEI values during periods when they remain continuously below −1 [3].

2.3.5. The Cross-Wavelet Transform Method

The cross-wavelet transform is an advanced analytical technique used to explore the relationship between two time series across both time and frequency domains. This method builds upon wavelet transform principles to examine time-dependent interactions between the periodic components of two signals [47]. In this study, the cross-wavelet transform was applied to analyze the connections between SPI and SPEI, SPI and SRHI, and SPEI and SRHI across various regions of Benin. Comprehensive details on the theoretical framework of cross-wavelet power distribution between two time series are available in [37,48].

2.3.6. Standardized Relative Humidity Index (SRHI)

The SRHI serves as a standardized measure to evaluate how RH deviates from its historical climatological average. RH indicates the air’s moisture content relative to its maximum moisture-holding capacity at a specific temperature [21]. Expressed as a percentage, 0% denotes completely dry air, while 100% represents fully saturated air. RH is derived by comparing the current absolute humidity with the highest possible absolute humidity at a given temperature. By incorporating atmospheric moisture conditions, SRHI serves as a critical indicator for drought monitoring, providing a comprehensive understanding of drought dynamics [11,21]. The SRHI offers a standardized measure of RH, allowing for more consistent comparisons across different climatic regions and seasons. By contextualizing RH values within historical norms, this index enhances the detection of anomalous moisture conditions that may not be evident from RH measurements alone. A positive SRHI indicates above-average RH for a given time and location, whereas a negative SRHI signifies below-average RH. Similar to the SPI and SPEI, the SRHI can be computed over various timescales (e.g., 1, 3, 6, or 12 months), facilitating comparative analyses of moisture variability and drought conditions. In [15,21], a non-parametric approach was used to determine the SRHI. This indicator is adaptable to various climates because it does not presume a particular distribution for RH data. This indicator is adaptable to various climates because it does not presume a particular distribution for RH data. Thus, comparing the SRHI to these well-known indicators, such as the SPI and SPEI, encourages more accurate drought monitoring and management.

3. Results

3.1. Analysis of Temporal Variations in SPI and SPEI

The SPI and SPEI were calculated for three distinct climatic zones (Zones 1–3) and the entire territory of Benin over the period from 1988 to 2021. These indices were computed at four temporal scales—one month, three months, six months, and twelve months—to capture short- and long-term drought variability (see Figure 5).

In Zone 1, the short-term SPI identified extreme drought events in 1988, 1990, 1992, 1993, 1997, and 2017 (see Figure 5i,j), while long-term droughts were detected in 1990, 1993, 2008, and 2012 (see Figure 5k,l). Conversely, the SPEI classified the following years as experiencing short-term drought: 1988, 1990, 1991, 2001, 2006, 2008, 2011, 2015, 2017, and 2021, while long-term drought events were identified in 1988, 1991, 2006, 2010, and 2017. Notably, the SPEI demonstrated superior performance in detecting severe and extreme drought conditions across northern Benin over both short- and long-term periods. However, it was unable to capture certain extreme drought events.

In Zone 2, the SPI identified a greater number of drought episodes compared to the SPEI. Specifically, the short-term SPI detected drought events in 1989, 1992, 1993, 1997, 2001, 2005, 2013, 2015, 2017, 2018, and 2020, whereas the short-term SPEI only identified droughts in 2005, 2006, 2007, 2008, and 2017. Similarly, for long-term drought events, the SPI recorded occurrences in 1993, 2005, 2006, 2013, 2015, 2017, and 2018, while the long-term SPEI detected droughts in 2006 and 2007 (Figure 5e–h).

In Zone 3, the SPI identified extreme and exceptional short-term droughts (1–3 months) during the periods of 1990–1992, 1996–1998, 2001–2003, 2009, 2011–2013, 2015, 2017, 2020, and 2021 (Figure 5a,b). Additionally, the SPI detected drought events lasting between six and twelve months in 1990, 1992, 1993, 1997, 2000, 2002, 2013, 2015, 2016, and 2021 (Figure 5c,d). Conversely, the SPEI identified short-term drought episodes in 1992, 1993, 1996, 1998, 2000, 2001, 2007, 2009, 2016, 2017, 2020, and 2021, along with a limited number of long-term drought occurrences in 1992, 1993, 1998, 2001, 2002, 2013, and 2021. Overall, the SPI demonstrated superior performance in detecting severe and extreme drought events in Zone 3 across both short- and long-term timescales compared to the SPEI.

In terms of drought detection, SPEI performs better in Zone 1 due to its sensitivity to temperature-induced evapotranspiration, while SPI performs better in Zone 2, where precipitation is a more important factor. In Zone 3, SPI again proves superior, particularly in detecting frequent and intense short-term droughts.

The SPI identified a greater number of severe and extreme drought events across Benin compared to the SPEI. According to the SPI, severe and intense drought conditions occurred in 1992, 2001, 2015, 2017, and 2020 for short durations (Figure 6a,b), with a prolonged drought event recorded in 2015 (Figure 6c,d). In contrast, while the SPEI detected severe and exceptional drought conditions during short-term drought events, it failed to identify any long-term droughts in 2017.

Figure 7 shows the comparison of PET when SPI has only droughts, SPEI has only droughts, and both SPI and SPEI have droughts. Seasonal variations in PET reveal that PET is generally lower during the wet season and higher during the dry season in all three zones. In Zone 1, PET ranges from 102 to 147 mm/month during SPI droughts only, increases to 102 to 159 mm/month during SPEI droughts only, and reaches 105 to 151 mm/month when both SPI and SPEI indicate drought conditions (see Figure 7a–c). In Zone 2, PET ranges from 120 to 198 mm/month during SPI droughts only, increases to 133 to 200 mm/month during SPEI droughts only, and ranges from 128 to 202 mm/month during simultaneous SPI and SPEI droughts (see Figure 7d,e). In Zone 3, PET ranges from 123 to 199 mm/month during SPI droughts only, from 135 to 196 mm/month during SPEI droughts only, and further increases to 125 to 212 mm/month when both indices signaled drought conditions (see Figure 7f,g). Monthly PET values in each zone show significant variability, with higher values observed particularly during the dry season and lower values during the wet periods. This increase in PET during dry periods is largely due to high temperatures and reduced RH, which intensifies atmospheric water demand. Conversely, PET decreases during the wet season due to cooler temperatures and high RH. These seasonal fluctuations in PET highlight increased water demand during the warmer months, with implications for agricultural water management, particularly during the summer growing season when irrigation needs are typically higher.

The PET profiles between zones revealed regional differences: Zone 1 exhibits moderately high PET values, particularly during SPEI-only droughts; Zone 2 exhibits the widest range of PET values, with peaks reaching up to 202 mm/month during simultaneous SPI–SPEI droughts, indicating significant climate stress; and Zone 3 records the highest PET values, reaching up to 212 mm/month, highlighting a high demand for atmospheric water.

3.2. Man–Kendall Trend Analysis

The outcomes of the Mann–Kendall trend analysis and Sen’s Slope estimator for precipitation and temperature across the three climatic zones are displayed in

Table 1. MK trend test coefficient and significance coefficient (P) of precipitation and temperature for three different climate zones.

	Zones	Trend (Year−1)	Z	P	Sen’s Slope
Precipitation	Zone 1	0.09000614	1.39349767	0.16346935	0.0843
	Zone 2	−0.06952210	−0.5929777	0.55319605	−0.0838
	Zone 3	−0.03985643	−0.2964888	0.76685677	−0.079
Temperature	Zone 1	0.02339217	5.30715071	0.00000011	0.0124
	Zone 2	0.03364282	4.65487520	0.00000324	0.0326
	Zone 3	0.01186139	2.10507095	0.03528513	0.0235

. In Zone 1, precipitation exhibits an upward trend, whereas Zones 2 and 3 show a decline; however, these trends are not statistically significant (p > 0.05). Conversely, temperature demonstrates a significant upward trend (p < 0.05) across all three zones. Here, we see a difference in trend between the two variables (i.e., precipitation and temperature); this indicates some significant differences in the drought trends in SPI and SPEI values.

The findings of the Mann–Kendall trend test for the three climate zones and the entire country are summarized in

Table 2. MK trend test coefficients and significance levels (p-values) for SPI and SPEI across three distinct climatic zones and the entirety of Benin at different timescales.

	SPI	Trend (Year−1)	Z	P	SPEI	Trend (Year−1)	Z	P
Zone 1	SPI-1	0.000286	1.398	0.162	SPEI-1	−0.00105	−2.873	0.004
	SPI-3	0.000602	1.968	0.048	SPEI-3	−0.00085	−2.387	0.016
	SPI-6	0.000979	2.785	0.005	SPEI-6	−0.00014	−0.413	0.679
	SPI-12	0.001554	4.916	0.000	SPEI-12	0.000432	1.1987	0.230
Zone 2	SPI-1	−0.00016	−0.618	0.535	SPEI-1	−0.00104	−2.988	0.002
	SPI-3	−0.00040	−1.196	0.231	SPEI-3	−0.00117	−3.379	0.000
	SPI-6	−0.00056	−1.598	0.109	SPEI-6	−0.00132	−3.865	0.000
	SPI-12	−0.00112	−2.940	0.080	SPEI-12	−0.00188	−5.350	0.000
Zone 3	SPI-1	0.00003	0.216	0.888	SPEI-1	0.00005	0.142	0.887
	SPI-3	−0.00016	−0.426	0.669	SPEI-3	−0.00004	−0.116	0.907
	SPI-6	−0.00051	−1.508	0.131	SPEI-6	−0.00029	−0.841	0.399
	SPI-12	−0.00107	−3.081	0.002	SPEI-12	−0.00069	−2.017	0.043
Whole Benin	SPI-1	0.000224	0.933	0.350	SPEI-1	0.000224	0.9334	0.350
	SPI-3	0.000115	0.484	0.627	SPEI-3	−0.00073	−2.847	0.004
	SPI-6	−0.000146	−0.602	0.547	SPEI-6	−0.00065	−2.591	0.009
	SPI-12	−0.000241	−0.945	0.344	SPEI-12	−0.00080	−3.280	0.001

. In Zone 1, significant trends (p < 0.05) are detected across all timescales, except for SPI-1, which shows no significance. The SPI values at 1-, 3-, 6-, and 12-month intervals display an overall upward trend. In contrast, the SPEI in Zone 1 declines at 1-, 3-, and 6-month timescales, with SPEI-1 and SPEI-3 showing statistically significant decreases (p < 0.05), while SPEI-6 and SPEI-12 do not present significant trends. Notably, while SPEI-12 displays an upward trend, it is not statistically significant. In Zone 2, both SPI and SPEI exhibit a declining trend across all timescales. However, while the decreasing trend is statistically significant for SPEI (p < 0.05), it remains statistically insignificant for SPI (p > 0.05). In Zone 3, SPI-1 and SPEI-1 exhibit a slight upward trend, though minimal. However, both indices demonstrate decreasing trends at the 3-, 6-, and 12-month scales. While SPI-12 and SPEI-12 show a more noticeable decline, these trends, along with the others, remain statistically insignificant (p > 0.05). At the national level, SPI-1 and SPI-3 exhibit an increasing trend, whereas SPI-6 and SPI-12 show a downward trend, though neither is statistically significant (p > 0.05). In contrast, SPEI at the 3-, 6-, and 12-month timescales demonstrates a significant declining trend (p < 0.05), while SPEI-1 presents an insignificant increasing trend. Overall, the trend analysis suggests that drought tendencies decline with increasing timescales in Zones 2 and 3, as well as at the national level. However, trends in Zone 1 exhibit greater complexity, with a decreasing tendency over longer timescales, underscoring distinct regional variations in drought dynamics across Benin.

Climate trend analysis highlights the difference between Zones 1, 2, and 3: Zone 1 shows a non-significant increase in precipitation and a significant rise in temperature, while Zones 2 and 3 show a (non-significant) decrease in precipitation associated with significant warming, indicating increasing aridity.

The M–K test results across various areas reveal significant variations in the sensitivity of SPI and SPEI values across varying timescales. Since SPI calculations rely solely on precipitation data, their values are directly influenced by fluctuations in precipitation. In contrast, SPEI incorporates both precipitation and evapotranspiration, making it more responsive to changes in drought conditions. This inclusion of evapotranspiration enhances SPEI’s sensitivity to drought variability, resulting in observable differences from SPI.

A comparison of Table 1 and Table 2 reveals that, in Zones 2 and 3, the trends of SPI and SPEI are consistent, with both indices showing negative trends for SPI-3–6–12 and SPEI-3–6–12. This consistency is attributed to a decreasing trend in precipitation and an increasing trend in temperature in these zones.

The decline in precipitation directly causes a reduction in SPI values, while the combined effects of decreased precipitation and increased temperature contribute to the decline in SPEI values.

In Zone 1, the trends are distinct. While precipitation exhibits an increasing trend alongside a warming trend in temperature, the effects on the indices vary. For SPI-3–6, the increased precipitation leads to an upward trend. However, for SPEI-3–6, the increase in potential evapotranspiration due to higher temperatures results in a decline. Notably, both SPI-12 and SPEI-12 in Zone 1 show an increasing trend, indicating that, at the 12-month scale, the dominant influence of increased precipitation outweighs the impact of rising temperatures on the SPEI.

This study evaluated the correlation analysis between SPI and SPEI at different timescales by applying Pearson’s correlation coefficient. The findings revealed a robust and statistically significant link between SPI and SPEI over 1-, 3-, 6-, and 12-month intervals, with correlation values of r = 0.63, r = 0.94, r = 0.96, and r = 0.97, respectively. Interestingly, the relationship strengthened at longer timescales (6 and 12 months), where the coefficients reached r = 0.96 and r = 0.97, indicating that the linear connection between SPI and SPEI grows stronger with extended temporal aggregation.

3.3. Characteristics of Drought Events: Duration and Intensity

Table 3. Overview of the maximum drought duration (D) and greatest drought intensity (I) derived from SPI and SPEI in various climatic regions of Benin.

Zones	Times Scales	Longest		Highest		Times Scales	Longest		Highest
		Year	D	Year	I		Year	D	Year	I
Zone 1	SPI-1	1990	4	2020	−1.91	SPEI-1	1992 2001 2015	4	2020	−1.67
	SPI-3	2001	8	2012	−1.55	SPEI-3	1992 2001 2015	4	2020	−1.94
	SPI-6	2001	6	1993	−1.68	SPEI-6	2001	6	1997	−1.57
	SPI-12	2001	8	2021	−1.48	SPEI-12	2001	8	2001	−1.35
Zone 2	SPI-1	2015	4	2020	−2.53	SPEI-1	2005 2006 2008 2015	4	2020	−2.53
	SPI-3	2005	5	2001	−1.9	SPEI-3	2010	4	2001	−1.9
	SPI-6	2005	8	2015	−1.56	SPEI-6	2003	7	2015	−1.5
	SPI-12	2016	12	1993	−1.58	SPEI-12	2016	11	2006	−1.49
Zone 3	SPI-1	2015	3	1988	−2	SPEI-1	2010	4	1991	−1.86
	SPI-3	1990	4	2015	−2	SPEI-3	2006	6	2010	−1.90
	SPI-6	1988 1993	5	1998	−1.75	SPEI-6	2006	8	2017	−1.78
	SPI-12	1988 1989 1994	6	2005	−1.69	SPEI-12	1988	7	2015	−1.49

summarizes the results of drought duration (D) and intensity (I) calculations for Zones 1, 2, and 3 across 1-, 3-, 6-, and 12-month timescales. The findings reveal significant variations in drought characteristics among the three zones, demonstrating clear differences in both duration and intensity. Over the 34 years (1988–2021), the longest drought event was recorded in Zone 2, lasting 12 months with an intensity of −1.58 in 1993. In contrast, the shortest drought was observed in Zone 3, lasting 3 months with an intensity of −2.0.

Furthermore, across all three climatic zones, it was observed that drought duration tends to increase with longer timescales, whereas drought intensity tends to decrease. These results emphasize that longer droughts do not necessarily correspond to more severe drought conditions. Understanding the duration and intensity of droughts at different timescales contributes to agricultural planning and environmental monitoring in Benin.

3.4. Examining the Correlation Between SPI and SPEI Using Cross-Wavelet Transform Analysis

Figure 8 illustrates the application of cross-wavelet transform analysis to examine the relationship between SPI and SPEI across different climatic zones. In Zone 1, a strong positive association is observed, with significant signals detected during 1992–1999 and 2006–2019, spanning timescales of 4–32 months. Similarly, Zone 2 exhibits a pronounced positive correlation, with major resonance cycles occurring between 5 and 15 months during 1992–1993, 1996, 1997, 2012, and 2016, as well as an extended cycle of 50–65 months from 1995 to 2008. In Zone 3, the SPI and SPEI show a positive phase correlation during two periods: 1989–1994 (9–15 months) and 1996–2000 (16–30 months). Across all zones, the phase angle variations suggest a stable directional link between the two indices. At the national scale, the cross-wavelet analysis reveals two distinct periods of strong correlation: the first, encompassing 1990, 1996, and 2000, with cycles of 8–16 months, and the second, spanning 1997–2006, with cycles of 50–65 months. Additionally, the phase angle dynamics suggest that SPI exerts a stronger influence on SPEI fluctuations across Benin.

3.5. Investigating the Correlation of SPI with SRHI and SPEI with SRHI

The SRHI is also widely used for drought monitoring [10,13,20], and its relationship with the SPI and SPEI indices has been examined across the different zones in Benin. The strongest correlation coefficients between SRHI and SPI/SPEI were primarily observed between SRHI-1 and SPI-1/SPEI-1 (see

Table 4. The correlations between the SRHI and both the SPI and SPEI at 1-, 3-, 6-, and 12-month timescales for different climatic zones.

Zones	Relationship	Correlation Coefficient	Relationship	Correlation Coefficient
Zone 1	SPI-1 and SRHI-1	0.40	SPEI-1 and SRHI-1	0.25
	SPI-3 and SRHI-3	0.29	SPEI-3 and SRHI-3	0.15
	SPI-6 and SRHI-6	0.29	SPEI-6 and SRHI-6	0.25
	SPI-12 and SRHI-12	0.39	SPEI-12 and SRHI-12	0.36
Zone 2	SPI-1 and SRHI-1	0.34	SPEI-1 and SRHI-1	0.18
	SPI-3 and SRHI-3	0.27	SPEI-3 and SRHI-3	0.28
	SPI-6 and SRHI-6	0.13	SPEI-6 and SRHI-6	0.14
	SPI-12 and SRHI-12	0.07	SPEI-12 and SRHI-12	0.08
Zone 3	SPI-1 and SRHI-1	0.45	SPEI-1 and SRHI-1	0.55
	SPI-3 and SRHI-3	0.42	SPEI-3 and SRHI-3	0.52
	SPI-6 and SRHI-6	0.36	SPEI-6 and SRHI-6	0.44
	SPI-12 and SRHI-12	0.38	SPEI-12 and SRHI-12	0.41
Benin	SPI-1 and SRHI-1	0.41	SPEI-1 and SRHI-1	0.50
	SPI-3 and SRHI-3	0.48	SPEI-3 and SRHI-3	0.43
	SPI-6 and SRHI-6	0.38	SPEI-6 and SRHI-6	0.37
	SPI-12 and SRHI-12	0.30	SPEI-12 and SRHI-12	0.26

All coefficients are significant with p < 0.05.

). In Zones 1 and 2, the correlation between SRHI-1 and SPI-1 was the highest, with coefficients of r = 0.40 and r = 0.34, respectively. Conversely, in Zone 3 and across the whole of Benin, SRHI-1 showed a stronger correlation with SPEI-1, with coefficients of r = 0.55 and r = 0.50, respectively (Table 3). These results suggest that SRHI has a closer relationship with SPI than with SPEI in Zones 1 and 2, whereas SRHI correlates more strongly with SPEI than with SPI in Zone 3 and at the national scale.

4. Discussion

Different drought indices (such as SPI, SPEI, and SRHI) have been used by various researchers to study the drought vulnerability of Benin [17,49,50]. For instance, [12,35] showed that it is quite difficult to quantify the impact of PET on drought conditions. Unlike in previous works [17,49,51] that studied Benin’s drought vulnerability, this study uses SPI and SPEI along with SRHI to study the region’s drought. The present study’s adoption of SRHI can further help improve the detection of the onset of drought and assist in early warning systems.

The analysis of temporal variations and trend in the two precipitation-based indices supports numerous studies across the climatic zones in Benin [17,37,50], which revealed that there has been a rainfall deficit with an increase in temperature over the past five decades. The results from this study are consistent with the works in [16,28,52], who showed severe and extensive drought over the past five decades throughout West Africa. Furthermore, an analysis of obtained results shows that, since the year 2000, there have been deficits in rainfall, which indicate that drought conditions in the study region are still not over. This observation is also consistent with the findings in [44,49]. Elsewhere in other parts of Africa and in China’s Ningxia and Shaanxi provinces and the Yangtze River Basin, similar observations have been made [23,52]. This suggests that this could have a global extent.

The analysis of SPI and SPEI over time in Benin’s various climatic zones offered key insights into drought patterns and trends between 1988 and 2021. Both indices showed higher consistency over longer time periods, which supports earlier research indicating that extended timeframes improve the accuracy of drought evaluations [17,49]. The trend analysis revealed significant spatial variability in drought trends. In Zone 1, the SPI demonstrated a significant increasing trend for the 3-, 6-, and 12-month timescales, whereas the SPEI exhibited a significant decreasing trend for the 1- and 3-month timescales. This discrepancy emphasizes how crucial it is to take temperature and precipitation into account when evaluating drought [35]. Both SPI and SPEI in Zones 2 and 3 showed a slow decline in drought trends over longer durations, highlighting regional differences in how the climate reacts to drought.

The longest drought duration and highest drought intensity were identified for each zone. The longest droughts at shorter timescales (SPI-1, SPI-3) and (SPEI-1, SPEI-3) occurred in Zone 1 and Zone 3, respectively, while the most intense droughts at these timescales were recorded in Zone 2. In contrast, the longest droughts at longer timescales (SPI-6, SPI-12, SPEI-6, and SPEI-12) were observed in Zone 2, while the most intense droughts at these timescales occurred in Zone 3. These results provide insight into the spatiotemporal characteristics of drought events, highlighting variations in drought behavior across climatic zones.

By applying cross-wavelet transform analysis, the study clarified the link between SPI and SPEI across various zones, demonstrating a consistent positive relationship and significant recurring cycles at different intervals. In Zone 1, the consistency of drought signals among the indices was reflected in the strong correlations between SPI and SPEI over 4- to 32-month periods from 1992–1999 and 2006–2019. Likewise, in Zone 3, marked positive correlations appeared in the 9- to 15-month range (1989–1994) and the 16- to 30-month range (1996–2000). These findings point to a continuous directional link between SPI and SPEI, suggesting that, in some areas, SPI can act as a precursor to SPEI and vice versa.

The correlations between the SRHI and the SPI, as well as between the SPEI and SRHI, were analyzed to evaluate the sensitivity of these drought indices to relative humidity variations. This evaluation was carried out to determine the best drought-monitoring index for Benin’s varied climatic regions. Both SPI and SPEI exhibited strong positive correlations with SRHI at all timescales, with the highest correlations observed between SRHI-1–3/SPEI-1–3 and SRHI-1–3/SPI-1–3 in the different zones. This result highlights the greater sensitivity of SPI and SPEI to variations in precipitation, temperature and RH. The robust relationship between SRHI/SPI at the 1–3-month timescale in Zones 1–2 and SRHI/SPEI in Zone 3 highlights the utility of SPI for drought assessment in Zones 1–2 and SPEI in Zone 3. However, further research is required because of the regional variations in drought trends and the constraints brought on by data uncertainties.

The findings of this study are consistent with historical drought patterns documented in Benin and West Africa. According to the Intergovernmental Panel on Climate Change (IPCC 2018) [53], both average and extreme temperatures have increased in West Africa due to climate change. In Benin, [37] identified a declining trend in precipitation across the three climatic zones, with Zone 3 experiencing a sharp decrease in rainy days (from 128 days per year in 1960 to 80 in 2008) and Zone 1 showing a decline in annual rainfall from 1220 mm in 1962 to 1100 mm in 2008. These climatic shifts were accompanied by an increase in temperature—notably, a rise of more than 1 °C across zones from 1960 to 2008, with Zone 2 exhibiting a rate of 0.03 °C per year [37]. Table 1 in the present study similarly highlights a decrease in rainfall in Zones 2 and 3, alongside increasing temperature trends. Zone 3, in particular, experienced severe and extreme drought events during 1990–1995 and 2000–2005, indicating its heightened sensitivity to climatic stressors [19]. Other studies [28,54] have similarly emphasized Benin’s vulnerability to global warming. Notably, significant drought years such as 1983, 1992, 1995, and 2010 were consistent with broader regional patterns [4]. Similar drought conditions have been reported in the Upper Ouémé River Basin, the Komadugu–Yobe Basin, and the Sahel region [49,54,55].

Overall, both the SPI and SPEI indices used in this study consistently detect drought across all evaluated timescales, reinforcing the reliability of the findings. The use of recent observational data from the National Meteorological Agency of Benin (NMAB) offers a timely update relative to earlier studies. These insights are crucial for policymakers aiming to manage severe drought risks, address water scarcity challenges, and implement targeted strategies to mitigate the socio-economic and environmental impacts of drought in Benin.

5. Conclusions

This study analyzed meteorological drought characteristics across three distinct climatic zones in Benin by employing SPI and SPEI indices. A comprehensive methodological approach was utilized, incorporating the Mann–Kendall trend test, Pearson correlation analysis, drought duration and intensity assessment, cross-wavelet transform analysis, and evaluation of SRHI relationships. The findings reveal increasing consistency in SPI and SPEI trends over longer timescales, with significant regional variations. Zone 1 exhibited an increasing SPI trend across all timescales, whereas SPEI showed a decreasing trend at shorter timescales. In Zones 2 and 3, both indices generally indicated a decreasing trend, except at the one-month scale. Drought duration and intensity analysis revealed that the longest droughts at shorter timescales (SPI-1, SPI-3, SPEI-1, SPEI-3) occurred in Zones 1 and 3, while the most intense events were recorded in Zone 2. At longer timescales (SPI-6, SPI-12, SPEI-6, SPEI-12), Zone 2 experienced the longest droughts, whereas Zone 3 recorded the highest intensities. These variations highlight the complexity of drought patterns and emphasize the need for region-specific monitoring strategies. Cross-wavelet analysis demonstrated a strong correlation between SPI and SPEI, particularly at 6- and 12-month intervals. Additionally, SRHI exhibited a stronger correlation with SPI than SPEI in Zones 1 and 2, while, in Zone 3 and across Benin as a whole, SRHI correlated more strongly with SPEI. These interzonal differences highlight the importance of selecting drought indices and monitoring strategies that are tailored to the specific climate dynamics of each region. In terms of drought detection, SPEI performs better in Zone 1 due to its sensitivity to temperature-induced evapotranspiration, while SPI proves more effective in Zone 2, where precipitation is the dominant factor. In Zone 3, SPI again demonstrates superior performance, particularly in detecting frequent and intense short-term droughts. These recommendations aim to support the development of more effective drought early warning systems and agricultural planning frameworks adapted to local conditions. It is also important to acknowledge the limitations related to environmental diversity. Future studies should incorporate more diverse datasets and explore additional drought indicators to strengthen policy-making and advance drought resilience strategies for agriculture, ecosystems, and food security in Benin.