**Table 1.**

*Meteorological stations used.*

#### **3.2 Data analysis**

The rainfall amounts for months with effective rainfall for plant growth (April to October) for the nine (9) chosen stations were analysed with the Bhalme and Mooley Drought Index (BMDI) to determine drought intensity [48]. For farming activities, the growing season from April to October is taken to be very crucial in drought research and assessments. This is so, as these months are when about 95% of the entire annual rainfall is received in the study area [49].

Generally, the BMDI for a particular month K is determined with the formula:

$$\mathbf{I}\_{\mathbf{k}} = (\mathbf{M}\mathbf{k}/\mathbf{d}) + (\mathbf{1} + \mathbf{C})\,\mathbf{I}\_{\mathbf{k}-\mathbf{1}} \tag{1}$$

Where:

C is taken as a constant.

d is taken as aconstant.

IK = drought severity for the Kth month.

Ik-1 = drought severity for the (K-1) month.

M = moisture index is determined by.

$$\mathbf{M} = \mathbf{100} \ (\mathbf{x} \overline{-\mathbf{x}}) / \mathbf{S} \tag{2}$$

In Eq. (2),

X is the monthly rainfall amount; *x* is the long period average monthly rainfall; and. S is the standard deviation for the initial month being considered (K-1). Eq. (1) is therefore given as:

$$\mathbf{I} = \mathbf{M}/\mathbf{d} \tag{3}$$

The values of C and d in Eq. (3) for the Northern area in Nigeria are 0.43 and 38.84, respectively. These constants were derived by [48]. They were used in Eqs. (1) and (3) to determinethe monthly values of BMDI of the stations used. From the values, the averages or seasonal drought index (SDI) series were derived for each year in the study period. The seasonal indices were used to classify a year into any of the following wetness/dryness categories, using [48] classification chart (**Table 2**).

*Meteorological Drought and Temperature in Sudano-Sahelian Region of Nigeria… DOI: http://dx.doi.org/10.5772/intechopen.100108*


#### **Table 2.**

*BMDI classification chart.*

The rainfall data were divided into non-overlapping decades; 1961–1970, 1911– 1980, and so on to the present decadeof 2011–2020 and temperature decade of 1981–1990, 1991–2000 to the present decade of 2011–2020. Cramer's test as given by [50], was utilized in the compiling of the means of the decades with the mean of the entire study period. In the application of Cramer's test, the mean (*x*), and the standard deviation (*δ*), were determined for the stations and the total number of years under study, *N*. As mentioned earlier, this was to find out the difference (in terms of a moving *t*-statistic) between the mean (*xk*), for each successive n-year period and the mean (*x*) for the entire study period. The *t*-statistic was taken as:

$$t\_k = \left(\frac{n(N-2)}{N-n\left(1+\tau\_k^2\right)}\right)^{1/2}\tau\_k\tag{4}$$

with *τ<sup>k</sup>* as a standardized measure of the difference between means given as:

$$
\pi\_k = \frac{\overline{\boldsymbol{x}}k - \overline{\boldsymbol{x}}}{\delta} \tag{5}
$$

with *xk* as the mean of the sub-period of n-years. *x* and *δ* as the mean and standard deviation of the entire series respectively and *τ<sup>k</sup>* as the value of the student *t*-distribution with *N*-2 degrees of freedom. It was then compared against the "students" *t*-distribution table, at a 95% confidence level as regards the two-tailed form test. When *tk* was outside the bounds of the two-tailed probability of the Gaussian distribution (equal to 1.96 at 95% confidence level), a significant shift from the mean was assumed.

The temperature series were divided into 30-year overlapping sub-periods of 1981–2010 and 1991–2020, while the rainfall series were divided into 30-year nonoverlapping sub-periods of 1961–1990 and 1991–2020 as recommended by the World Meteorological Organisation. The student's *t*-test, *td*, was later used to ascertain if the sub-period means have deviated significantly in the occurrence of wet and dry years through time. The statistics, *td*, was calculated using:

$$t\_d = \frac{\left(\overline{X}\_2 - \overline{X}\_1\right) - \left(\mu\_2 - \mu\_1\right)}{\left[\frac{N\_2 S\_2^2 + N\_1 S\_1^2}{N\_2 + N\_1 - 2} \cdot \frac{1}{N\_2} + \frac{1}{N\_1}\right]^{\frac{1}{\lambda\_2}}}\tag{6}$$

with (*X*1–*X*2) being the difference in group means, (*μ*2–*μ*1), was the expected differences (set equal to 0), *N*<sup>2</sup> and *N*<sup>1</sup> being the number of cases in each subsample, and *S*<sup>2</sup> and *S*<sup>1</sup> were the respective standard deviations. When *td* was outside the bounds of the two-tailed probability of the Gaussian distribution, equal to 1.96 at 95% confidence level, a significant deviation and shift from the mean were assumed.

### **4. Results and discussion**

#### **4.1 Frequency distribution of drought series**

The results of the analysis of the occurrence of drought, in the study area, using BMDI are depicted in **Figures 2**–**10**. They show that the occurrences and intensity of drought in sub-areas of the study area vary over time and space. This is exemplified by some sub-areas with intermittent years of drought and wetness. Yelwa subarea experienced such a situation, even though droughts of mild and moderate intensities prevailed between 1965 and 1991. The later years of study, 2010–2020 also experienced a drought of mild intensity. Potiskum sub-area values also indicated years of drought and wetness throughout the study period with intensity varying between mild and moderate and a reduction in the drought intensity towards the end of the study period. This same situation also occurred in Gusau and Kaduna sub-areas. For Kaduna, the years between 1961 and 2011 were those of alternating wet and drought with intensity being mild.

Some sub-areas experienced very clear and distinct periods of drought. Generally, Maiduguri drought for example can be categorized into before the 1980s with

**Figure 2.** *Annual fluctuations in Bhalme and Mooley drought index for Yelwa.*

**Figure 3.** *Annual fluctuations in Bhalme and Mooley drought index for Potiskum.*

*Meteorological Drought and Temperature in Sudano-Sahelian Region of Nigeria… DOI: http://dx.doi.org/10.5772/intechopen.100108*

**Figure 4.** *Annual fluctuations in Bhalme and Mooley drought index for Maiduguri.*

**Figure 5.** *Annual fluctuations in Bhalme and Mooley drought index for Gusau.*

**Figure 6.** *Annual fluctuations in Bhalme and Mooley drought index for Kano.*

drought years and intensity that were less than moderate. This was followed by the drought of the 1980s with an intensity of between mild and moderate and the post-1980s with intermittent drought and wet years and reduced drought intensity of less than moderate. In Kano, drought dominated from the beginning of the study period to 1995, while for Sokoto and Katsina the drought years were 1967 to 1995 and 1967 to 2000 respectively with the intensity ranging from mild to moderate. Nguru also experienced similar droughts situation between 1969 and 2015.

The occurrences of drought as discussed above and wet years in the study area as exemplified in Kano between 1996 and 2010, in Sokoto and Katsina after 1995 and

## *The Nature, Causes, Effects and Mitigation of Climate Change on the Environment*

**Figure 7.** *Annual fluctuations in Bhalme and Mooley drought index for Sokoto.*

#### **Figure 8.** *Annual fluctuations in Bhalme and Mooley drought index for Katsina.*

#### **Figure 9.**

*Annual fluctuations in Bhalme and Mooley drought index for Nguru.*

2000 respectively (with few drought years and reduce intensities), before 1969 and after 2015 in Nguru and between 2011 and 2020 in Kaduna indicate a constant shift in the climate of the study area. This shift from findings seems more temporary than permanent and therefore, indicates climate variability than climate change. This variability in climate in more recent years has been exemplified in the Kaduna sub-area that before 2013 had indicated wet years in earlier studies [19, 50] had by 2013 to 2020 turned to drought. This however is in agreement with other studies that indicated the return of drought in the sub-area [14, 17, 27]. This notion of

*Meteorological Drought and Temperature in Sudano-Sahelian Region of Nigeria… DOI: http://dx.doi.org/10.5772/intechopen.100108*

**Figure 10.** *Annual fluctuations in Bhalme and Mooley drought index for Kaduna.*

climate variability is further supported by the fact that throughout the study period, the intensity of drought fluctuated between mild and moderate and never increased to severe or extreme.
