Spatiotemporal Characteristics and Trends of Meteorological Droughts in the Wadi Mina Basin, Northwest Algeria

: Drought has become a recurrent phenomenon in Algeria in the last few decades. Signiﬁcant drought conditions were observed during the late 1980s and late 1990s. The agricultural sector and water resources have been under severe constraints from the recurrent droughts. In this study, spatial and temporal dimensions of meteorological droughts in the Wadi Mina basin (4900 km 2 ) were investigated to assess vulnerability. The Standardized Precipitation Index (SPI) method and GIS were used to detail temporal and geographical variations in drought based on monthly records for the period 1970–2010 at 16 rainfall stations located in the Wadi Mina basin. Trends in annual SPI for stations in the basin were analyzed using the Mann–Kendall test and Sen’s slope estimator. Results showed that the SPI was able to detect historical droughts in 1982/83, 1983/84, 1989/90, 1992/93, 1993/94, 1996/97, 1998/99, 1999/00, 2004/05 and 2006/07. Wet years were observed in 1971/72, 1972/73, 1995/96, 2008/09 and 2009/10. Six out of 16 stations had signiﬁcant decreasing precipitation trends (at 95% conﬁdence), whereas no stations had signiﬁcant increasing precipitation trends. Based on these ﬁndings, measures to ameliorate and mitigate the effects of droughts, especially the dominant intensity types, on the people, community and environment are suggested.


Introduction
Drought is a recurring phenomenon that affects a wide variety of sectors, making it difficult to develop a single definition of it. According to a water-resource-oriented definition, which takes into account the water requirements related to biological, economic and social characteristics of a region, drought refers to a condition of severe reduction of water supply availability (compared to a normal value), extending along a significant period of time over a large region [1]. Drought is a complex phenomenon that involves different human and natural factors that determine the risk and vulnerability to it [2].
The particularly strong influence of drought on many sectors is visible in arid and semiarid regions, where water is scarce [3]. Water scarcity can strongly impact the agricultural sector in such regions [4]. In the case of the Mediterranean Basin, much of which is arid or semiarid, the extremely variable precipitation across temporal and spatial scales is influenced by geographical position of the region between two contrasting masses of water: the Atlantic Ocean and the Mediterranean Sea [5][6][7]. An additional feature determining high variability of precipitation in this region is the presence of various mountain ranges distributed along the coastal areas from east to west [7]. To avoid water scarcity, increased Moreover, the water resources have high variability and projections are that rainfall could decrease by more than 20% by 2050, which would result in greatly worsening water shortages in different basins of Algeria [51]. Knowledge on extreme dry conditions is also very important because these can influence not only on water scarcity for agriculture but also on natural ecosystems, mainly forest in the case of the studied basin. For example, Mensah et al. [52] showed that elevated temperatures will further exacerbate the drought impacts on forest ecosystems at sites with precipitation levels equal or smaller than the atmospheric evaporative demand and strong influence of vapor pressure deficits on carbon uptake, and can worsen the decline in soil moisture.
In this work, the objective was to better characterize annual-scale drought patterns over the Wadi Mina basin in order to aid water resource planning. The main objectives are to (1) map characteristics of drought patterns over the basin during 1970-2010, (2) identify any trends in precipitation or in drought characteristics, (3) identify drought years over the observation period, and (4) estimate the return periods for severe drought across the basin.

Study Area
The Wadi Mina basin, with an area of 4900 km², is located in the northwest of Algeria ( Figure 1). The Wadi Mina involves four major tributaries: Wadi Mina, Wadi Haddad, Wadi Abd and Wadi Taht. The climate is continental, with cold winters and hot summers. Mean annual precipitation ranges from about 220 to 400 mm, and most precipitation occurs between November and March. Mean annual temperatures are about 16 °C to 19.5 °C. Almost half the basin is covered by a varying density of vegetation, with in particular 32% of scrub, 35.8% of forests and 20% cereal crops [53].

Data Used
Monthly precipitation records for a 40-year observation period (September 1970 to August 2010, using water years that go from September to August) are compiled for 16 stations from the Algeria National Agency of Water Resources ( Figure 1 and Table 1). These stations constitute a relatively well-distributed network with acceptable spatial density over the basin. To assure quality, data was checked for inhomogeneities using the double mass curve, linear regression and Mann-Whitney test methods. The procedure detected a few inhomogeneities, for which the irregular data were adjusted using data of nearby reliable stations. Rainfall data of these 16 stations were analyzed statistically to evaluate rainfall variability in the study area ( Table 2). These preliminary statistical analyses included measure of central tendency (mean, and median), dispersion (standard deviation SD, coefficient of variation CV) and distribution (skewness Cs and kurtosis Ck) ( Table 3). To test for stationarity, the Kwiatkowski-Phillips-Schmidt-Shin test (KPSS) was used [54]. The results of the stationarity test for monthly, seasonal and yearly series are shown in Table 3. All the monthly, seasonal and annual series of the rainfall stations are indicated as showing stationarity (p-value is more than 0.05).

SPI
The standardized precipitation index (SPI) is commonly used to detect meteorological drought. Each drought is characterized by drought intensity (Di), a drought magnitude (Dm) and drought duration (Dd). Run intensity can be either the value of the SPI at any moment (Dint) or the minimum SPI value during a drought event (Dmi) The drought magnitude (Dm) is equal to the accumulated values of below-threshold SPI during each drought event ( Figure 2 To test for stationarity, the Kwiatkowski-Phillips-Schmidt-Shin test (KPSS) was used [54]. The results of the stationarity test for monthly, seasonal and yearly series are shown in Table 3. All the monthly, seasonal and annual series of the rainfall stations are indicated as showing stationarity (p-value is more than 0.05).

SPI
The standardized precipitation index (SPI) is commonly used to detect meteorological drought. Each drought is characterized by drought intensity (Di), a drought magnitude (Dm) and drought duration (Dd). Run intensity can be either the value of the SPI at any moment (Dint) or the minimum SPI value during a drought event (Dmi) The drought magnitude (Dm) is equal to the accumulated values of below-threshold SPI during each drought event ( Figure 2).  SPI is mathematically based on the cumulative probability of monthly precipitation amount recorded at the observation post [56,57]. No evaporation estimate is considered, unlike other drought indices such as SPEI. SPI = 0 denotes average (climatological) precipitation, SPI = 1 denotes 1 standard deviation wetter than average, and SPI = −1 denotes 1 standard deviation drier than average. In the case of the presented analysis, the monthly precipitations were aggregated over water years, and finally a yearly SPI (12-month Water 2021, 13, 3103 6 of 26 timescale) for each water year was calculated. SPI periods (years) with SPI below the defined threshold are considered drought years, and consecutive drought years are grouped into droughts. The whole period of observation at a meteorological station is used to determine the parameters of a precipitation probability density function, taken to be in the form of a gamma distribution: where α and β are the shape and scale parameters respectively. x is consecutively precipitation and Γ(α) is the gamma function. The gamma function defined by the following: The alpha and beta parameters of the gamma distribution are estimated from the precipitation time series as where x is the mean value of precipitation quantity; n is the precipitation measurement number; x i is the quantity of precipitation in a sequence of data. The cumulative probability can be presented as: To allow for the possibility that the precipitation may be zero, a mixture probability distribution is used, for which the cumulative probability becomes where q is the probability that the quantity of precipitation equals zero. The calculation of the SPI is presented on the basis of the following equation [20,58]: where t is determined as and c 0 . c 1 . c 2 . d 1 . d 2 and d 3 are coefficients whose values are: According to McKee et al. [18] different categories and approximate probabilities of wet and dry spells can be considered based on SPI for the timescale of interest, as shown in Table 4. SPI is expected to follow a near-normal (bell curve) distribution, with SPI values Water 2021, 13, 3103 7 of 26 near 0 being the most common and high positive or negative SPI (corresponding to very wet or very dry periods, respectively) being rare. These probabilities shown in Table 4 are estimates, assuming that SPI is normally distributed. Achieving an approximately standard normal probability distribution is the main motivation behind the transformation of precipitation to SPI.

Trend Analysis
Trend analysis determines whether the measured values of a variable show a consistent increase or decrease during a time period. Many statistical methods can be used for trend detection in a time series of meteorological and hydrological records. In this study we used simple and accepted methods for evaluating trends, the Mann-Kendall test and Sen's estimator of slope.
The Mann-Kendall method is a widely used non-parametric test for detecting trends in climatological and hydrological time series. It has been suggested by many authors to assess trends in environmental data time series because, unlike least-squares linear regression, it is robust to outlying and extreme values.
The Mann-Kendall test statistic S is given by [59]: where n is the number of data. x are the data values at times j and k (j > k) and the sign function is The variance of S is computed by where t i is the number of ties of extent i and m is the number of tied rank groups. For n larger than 10, a Z test statistic that, under the null hypothesis of no correlation, approximates a standard normal distribution is computed as the Mann-Kendall test statistic as follows: If a linear trend is present in a time series, then the true slope (change per unit time) can be estimated by using a simple non-parametric procedure developed by Sen [60]. The slope estimates of the n(n − 1)/2 unique pairs of data are first computed by: where x j and x i are data values at time j and i (j > i). respectively. The median of these N values of Q is Sen's estimator of slope. After sorting the Q values, if N is even, then Sen's estimator is calculated by: If N is odd, then Sen's estimator is computed by: Sen's estimator Q med provides the rate of change and enables determination of the total change in any variable during the analysis period. Sen's slopes are expressed here as rate of change per 40 years (1970-2010) in mm.

Frequency Analysis
Drought frequency (F i ) is the chance of a station being in drought in a given year. This was estimated empirically based on the following formula: where n-number of years of drought (SPI equal 0 or less), N-number of analyzed years.

Drought Intensity (DI)
Drought intensity (DI) is used to represent the severity of the drought. The drought intensity of a site within a certain period is usually reflected by the SPI value. The more negative the SPI value, the more serious the drought is. Its formula is as follows:

Drought Magnitude (DM)
DM corresponds to the cumulative water deficit over a drought period. DM is the sum of the absolute values of all SPI values (0 or less) during a drought event (Equation (16)):

Drought Duration (DD)
DD equals the number of time periods between the drought start and its end. In our case, we consider all SPI values below 0 as drought years.

Return Period of Drought
In addition to computing drought frequencies as empirical probabilities in the 40-year observation record, return periods of severe drought were also computed in this study using the annual maximum series (AMS) approach. The AMS here is based on the time series of SPI values for drought years. A drought was described as an SPI value less than zero. Drought-free years were given a zero value. The number of years for which SPI values are available is used to calculate the duration of the sequence. Only non-zero values were used in the drought frequency calculation. To account for the number of zero values, a correction was made using nonexceedance likelihood (F ) according to the following expression [55,61]: where F is the non exceedance probability value obtained by using frequency analysis on the non zero values and q is the probability of zero values which can be calculated as the ratio of the number of time intervals without drought occurrences to the total number of time intervals in the recording period [55,61].
To estimate the return period of drought severity that may go beyond the values observed over the 40-year period for which we have data, we fitted a probability distribution to the derived AMS. In this case, the drought event time series were fitted with gamma distributions. The return period of drought with particular severity was then calculated as:

Temporal Variability
The SPI was used to provide an indicator of drought severity in this study. The temporal characteristics of droughts in Wadi Mina basin was analyzed based on the 12-month timescale water-year SPI computed for each station (Figure 3). Analysis of the computed SPI series shows the basin has experienced droughts of high severity and duration in the 1980s and 1990s. A drought is defined whenever the SPI reaches a value of 0.00 and continues until the SPI becomes positive again.
The  Table 2). Increase of trend of SPI and likely decreased intensity of drought is observed on three rain gauge stations located mainly in upper part of the basin, in the Wadi Abd tributary. Spatio-temporal changes of SPI is caused by change of precipitation. Elouissi et al. [62] detected similar decreasing trends of precipitation in the northern part of the Macta basin (Algeria), close to the Mediterranean coast, and increasing trends in the southern part. The changes of precipitation and SPI can be affected by geographical position of the area in relation to the Atlantic Ocean, the Mediterranean Sea and the Atlas mountain ranges [63]. We can also see from Figure 3 that dry periods have tendency to cluster over long stretches of years. Clustering is especially visible in station S8 during 1975-1993, S5 (1981-1999), S6 (1981-1999) and S13 (1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007). Figure 3 also shows that at station S3 located in the upper part of the Wadi Taht subbasin, and S6 and S5 in the upper Wadi Mina, intensity of meteorological drought since 2000, expressed by SPI, was small, with wet years being more common.
Trend analysis determines whether the measured values of a variable show a significant increase or decrease during a time period. In this study, we used a simple method for evaluating trends, Mann-Kendall test and annual and seasonal Sen's slopes of trend values are expressed as rate of change per 40 years (1970-2010) in mm. The result of this analysis is shown in Table 5. At 7 of 16 rain gauge stations, or 44% all stations, there was a significant negative trend (p < 0.1). The significance level of trend in 6 cases was p < 0.05, and in the case of S13 station the p value was under 0.01. At most stations, from 1996 onward there were mainly severe droughts. Significant decreasing trends were observed at stations

Spatial Variability
To visualize the distribution of droughts in the basin, the study area is divided using Theissen Polygon tool in Arc GIS 10.2 into 16 polygons corresponding to the 16 rainfall stations. Stations that are closely spaced are assigned less area and vice versa (Figure 4). Lee et al. [64] showed that the spatial distribution of the rain gauge networks and the density have a significant influence on accurately calculating areal precipitation and Thiessen method gave good results when the spatial distribution of the rain gauge networks was even, as was the case here. Moreover, the weights assigned to the different stations do not vary with time, and thus it is easy to map the precipitation falling during each period. Geostatistical methods offer more sophisticated approaches to making maps based on station data, but the uncertainty of areal precipitation is in any case high if there are relatively few stations, like in this basin [65]. Even though some stations may show drought conditions, a regional drought is acknowledged only when some major portion of the total study area is under drought. Regional drought is determined by the intra-annual precipitation distribution, which can be affected by teleconnection patterns [66,67], and the North Atlantic Oscillation indices [68]. Moreover, regional-scale influence on the rainfall conditions in North Africa could result from the response of the African summer monsoon to oceanic forcing, amplified by land-atmosphere interaction [69].
The spatial distribution of drought intensity is shown (Figure 4) Table 4. Number in branches above figures is number of station.  Table 1 presents drought classification for the 16 rain gauge stations in each year. The most common SPI category overall was near normal (NN). For several years (1971, 1972, 1995, 2008 and 2009) most stations were in wet categories (EW, VW and MW). For 1971, 1995 and 2008 only 2-5 out of 16 stations were dry, and no severe or extreme drought was observed. The highest number of stations with severe or extreme drought (SD and ED) was observed in the years 1981, 1983, 1989, 1992, 1996, 1998, 1999 and 2004. The highest number of years with unusually wet conditions (MW, VW and EW) were observed on stations S13 and S15-8 cases. These stations were located in the lower part of the Wadi Mina. The most cases of intense drought (ED, SD and ED) were observed at station S9-9 cases, and the highest number of years with severe and extreme drought were observed at stations S1, S9 and S12-4 cases.

Spatial Variability
To visualize the distribution of droughts in the basin, the study area is divided using Theissen Polygon tool in Arc GIS 10.2 into 16 polygons corresponding to the 16 rainfall stations. Stations that are closely spaced are assigned less area and vice versa (Figure 4). Lee et al. [64] showed that the spatial distribution of the rain gauge networks and the den-sity have a significant influence on accurately calculating areal precipitation and Thiessen method gave good results when the spatial distribution of the rain gauge networks was even, as was the case here. Moreover, the weights assigned to the different stations do not vary with time, and thus it is easy to map the precipitation falling during each period. Geostatistical methods offer more sophisticated approaches to making maps based on station data, but the uncertainty of areal precipitation is in any case high if there are relatively few stations, like in this basin [65]. Even though some stations may show drought conditions, a regional drought is acknowledged only when some major portion of the total study area is under drought. Regional drought is determined by the intra-annual precipitation distribution, which can be affected by teleconnection patterns [66,67], and the North Atlantic Oscillation indices [68]. Moreover, regional-scale influence on the rainfall conditions in North Africa could result from the response of the African summer monsoon to oceanic forcing, amplified by land-atmosphere interaction [69].
The spatial distribution of drought intensity is shown (Figure 4)

Frequency Analysis
Drought frequency calculated for all analyzed stations is presented in Table 6. Near normal (NN) conditions occurred most frequently at all stations (57.5% to 72.5% of the time, depending on station). The extreme categories-extreme wetness (EW) and extreme drought (ED)-were the least frequently observed. Extreme drought only occurred at 3 of 16 rainfall gauge stations over the 40-year observation period. However, all but 2 rainfall gauge stations (S4 and S8) observed either ED and SD. Table 2 shows drought duration, magnitude and intensity, as well as average, maximum and minimum SPI at annual time scale for the meteorological stations considered in this study area. All drought indicators have strong variability over the study area. Drought duration (DD) varied between 1 to 16 years, and most frequently was only 1 year. The highest DD of 12, 13 and 16 years were observed at stations S13, S6 and S5, respectively. The highest drought intensity was observed at S16 station and lowest at S8. The largest drought magnitude was observed at S5 and the smallest at S11. Extreme drought was observed in 1980/81, 1996/97 and 2004/05 in stations located in the lower part of basin. Table 6. Frequency of each drought and wetness class for the considered stations, %.

Frequency Analysis
Drought frequency calculated for all analyzed stations is presented in Table 6. Near normal (NN) conditions occurred most frequently at all stations (57.5% to 72.5% of the time, depending on station). The extreme categories-extreme wetness (EW) and extreme drought (ED)-were the least frequently observed. Extreme drought only occurred at 3 of 16 rainfall gauge stations over the 40-year observation period. However, all but 2 rainfall gauge stations (S4 and S8) observed either ED and SD. Table 6. Frequency of each drought and wetness class for the considered stations, %.

1
S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 0 2.  0  00  100  100  100  100  100  100  100  100  100  100  100  100  100  100  100   Table A2 shows drought duration, magnitude and intensity, as well as average, maximum and minimum SPI at annual time scale for the meteorological stations considered in this study area. All drought indicators have strong variability over the study area. Drought duration (DD) varied between 1 to 16 years, and most frequently was only 1 year. The highest DD of 12, 13 and 16 years were observed at stations S13, S6 and S5, respectively. The highest drought intensity was observed at S16 station and lowest at S8. The largest drought magnitude was observed at S5 and the smallest at S11. Extreme drought was observed in 1980/81, 1996/97 and 2004/05 in stations located in the lower part of basin.

Calculation of Return Period and the Severity-Area-Frequency (SAF) Curve
The Severity-Area-Frequency (SAF) curve is a very useful method for showing the spatial extent of different types of drought in a given area. This technique has been undertaken by several researchers around the world, for example in India [70]; China [71]; Southern Africa [72] and in Iraq [39]. Figure 5 illustrates the drought Severity-Area-Frequency curves of SPI annual scale time for 10-, 25-, 50-, 100-year exceedance

Calculation of Return Period and the Severity-Area-Frequency (SAF) Curve
The Severity-Area-Frequency (SAF) curve is a very useful method for showing the spatial extent of different types of drought in a given area. This technique has been undertaken by several researchers around the world, for example in India [70]; China [71]; Southern Africa [72] and in Iraq [39].  The severity analysis shows that all selected droughts have smaller severity than for return periods 10, 25, 50 and 100 years. Moreover, the high drought severity occurred on relatively small areas, less than 20% of analyzed basin and is observed mainly on the north part of the basin- Figure 4. Moderate or near normal years are observed on the most parts of the basin.

Spatial Pattern of Return Periods of Droughts
The return periods of moderate, severe and extreme droughts at all stations were calculated and the values were then used to prepare the corresponding maps by using inverse distance weighted (IDW) interpolation method analysis tool of ArcMap ( Figure  6). The presentation of these maps shows spatial variability of the drought for the different classes, with extreme drought more likely (shorter return period) in the north and east, apparently modulated by the high heterogeneity of the spatial distribution of the rainfall [45,73]. Assessing vulnerability to drought across the Wadi Mina basin is important, considering that, as shown by Henchiri et al. [74], grasslands and croplands in the The severity analysis shows that all selected droughts have smaller severity than for return periods 10, 25, 50 and 100 years. Moreover, the high drought severity occurred on relatively small areas, less than 20% of analyzed basin and is observed mainly on the north part of the basin- Figure 4. Moderate or near normal years are observed on the most parts of the basin.

Spatial Pattern of Return Periods of Droughts
The return periods of moderate, severe and extreme droughts at all stations were calculated and the values were then used to prepare the corresponding maps by using inverse distance weighted (IDW) interpolation method analysis tool of ArcMap ( Figure 6). The presentation of these maps shows spatial variability of the drought for the different classes, with extreme drought more likely (shorter return period) in the north and east, apparently modulated by the high heterogeneity of the spatial distribution of the rainfall [45,73]. Assessing vulnerability to drought across the Wadi Mina basin is important, considering that, as shown by Henchiri et al. [74], grasslands and croplands in the northern region of the Africa are highly vulnerable to drought risk.

Conclusions
This study was focused on analyzing temporal and spatial extents of droughts in the Wadi Mina basin, Algeria, using SPI as an indicator of drought severity. The aim of this study was to investigate spatial and temporal dimensions of meteorological droughts in the Wadi Mina basin. Meteorological drought was expressed by the Standardized Precipitation Index (SPI) method and GIS was used to detail temporal and geographical variations in the drought vulnerability based on severity of drought events at annual time steps. This study is applied to rainfall monthly records for the period 1970-2010 at 16 rainfall stations located in the Wadi Mina basin.
The results showed that the SPI was able to detect historical droughts of 1982/ 83, 1983/84, 1989/90, 1992/93, 1993/94, 1996/97, 1998/99, 1999 precipitation is already low. As expected given the process used to construct SPI, near normal conditions dominated at all stations, and severe and extreme drought categories were uncommon. The spatial variability of the drought showed that extreme drought is more likely (shorter return period) in the north and east.
Severity-Area-Frequency curves that can aid the development of a drought preparedness plan were developed for Wadi Mina basin, so as to ensure sustainable water resource planning within the basin.
One limitation of the study is that we used only SPI to detect drought intensities. In the future, we plan to add evapotranspiration and calculate the SPEI indicator, which can give more complex information about meteorological conditions influencing drought events, particularly for agricultural and forestry applications. Moreover, in this study only annual sum of precipitation was used and thus seasonal variability of drought was not detected. While this is to some extent justified for this region given that precipitation is concentrated in only a few months per year, in a future study monthly and seasonal To mitigate risks of drought, proper water management techniques must be adopted. One of these techniques is supplemental irrigation, which is an efficient practice used for increasing agricultural production under limited water resources in areas affected by drought [75].

Conclusions
This study was focused on analyzing temporal and spatial extents of droughts in the Wadi Mina basin, Algeria, using SPI as an indicator of drought severity. The aim of this study was to investigate spatial and temporal dimensions of meteorological droughts in the Wadi Mina basin. Meteorological drought was expressed by the Standardized Precipitation Index (SPI) method and GIS was used to detail temporal and geographical variations in the drought vulnerability based on severity of drought events at annual time steps. This study is applied to rainfall monthly records for the period 1970-2010 at 16 rainfall stations located in the Wadi Mina basin.
The results showed that the SPI was able to detect historical droughts of 1982/83, 1983/84, 1989/90, 1992/93, 1993/94, 1996/97, 1998/99, 1999 Most of the stations with the greatest decreasing trend were observed in the lower part of the Wadi Mina basin, where average precipitation is already low. As expected given the process used to construct SPI, near normal conditions dominated at all stations, and severe and extreme drought categories were uncommon. The spatial variability of the drought showed that extreme drought is more likely (shorter return period) in the north and east.
Severity-Area-Frequency curves that can aid the development of a drought preparedness plan were developed for Wadi Mina basin, so as to ensure sustainable water resource planning within the basin.
One limitation of the study is that we used only SPI to detect drought intensities. In the future, we plan to add evapotranspiration and calculate the SPEI indicator, which can give more complex information about meteorological conditions influencing drought events, particularly for agricultural and forestry applications. Moreover, in this study only annual sum of precipitation was used and thus seasonal variability of drought was not detected. While this is to some extent justified for this region given that precipitation is concentrated in only a few months per year, in a future study monthly and seasonal precipitation variations could also be explored. Moreover, as a future study, we plan to compare drought analyses based on different sources of rainfall data, including the Soil Moisture to Rainfall (SM2RAIN) [76] algorithm to estimate rainfall based on soil moisture time series.
The SM2RAIN is based on the inversion of the hydrological water balance, for estimation of rainfall from soil moisture observations. In this approach the soil is assumed as reservoir used for measuring the amount of rainfall [77]. This method gives independent rainfall product with a different error structure and allows integration with other satellite-based rainfall products. According to [76], the SM2RAIN method can be useful in regions for which satellite rainfall data are affected by higher errors or not available. Because Northwest Algeria is the region where water scarcity is high, we will perform analysis that can show potential use of SM2RAIN as indirect source of rainfall to detect meteorological drought, including seasonal variations.