Spatio-Statistical Analysis of Rainfall and Temperature Distribution, Anomaly and Trend in Nigeria

This study aims to examine the variability of rainfall and temperature based on spatiostatistical analysis. Datazforzthezstudy was gotten from the Nigerian Meteorological Agency and subjected to statistical analysis using mean, standard deviation, time series analysis, t-test and correlation. Thezresultszofzthezstudyzrevealedzthatzthe mean rainfall in the area is 108.6 mm, mean temperature is 28 C and mean sunshine is 4.7 hours. The result further revealed that mean onset date in the state is 13 march, mean cessation date is 10 October while thezmeanzlengthzofzrainyzseason is 223 days. The result also showed that rainfall anomaly index in the area ranged between -2.12 and 2.28 while temperature anomaly ranged between -2.31 and 1.73. The correlation coefficient showed that cessation (0.408) and Length of rainy season (0.406) is significantly related to rainfall, sunshine hours (0.380) and average temperature (0.867) is significantly related to minimum temperature, sunshine hours (-0.389) and average temperature (0.749) is significantly related to maximum temperature while onset (-0.642) and cessation (0.509) is significantly related to length of rainy season. However, there is a significant difference in onsetzdates,zcessationzdateszandzlengthzofzrainyzseasonzinzthezarea. The area is experiencing a significant increase in temperature, rainfall and sunshine hours and recommends that more tress should be planted in the area and Nigerian meteorological agency should also inform farmers about the onset of the rains so as to help the farmers prepare for the planting season.


Introduction
Rainfall is one of the major factors affecting food security especially in countries largely or highly dependent on rain-fed agriculture, given that, in addition to evaporation rate and soil characteristics, it controls the state of soil moisture. The role of moisture in agricultural production is even more important in the tropics, especially Nigeria, where rainfall is highly seasonal over most parts and varies from year to year, while the growing season is determined by the availability of rain to meet crop water requirements (Odekunle, 2004).
The amount of rainfall that is normally received determines which types of agriculture that can be carried out and which crops can be cultivated in a region. The seasonal distribution of rainfall regulates the agricultural calendar in the tropics (Vellinga et al., 2013). The relationship that exists between rainfall and the tropical occupation of agriculture in Nigeria is that it employs over 80% of the work force. Rainfall is arguably the most important meteorological parameter that has the greatest impact on human activity. Rainfall distribution and pattern has been a major concern to different people in diverse field. i.e. to the agriculturist, rainfall is a crucial factor that determines the planting season and influences the type of crops to be cultivated. To the hydrologist, rainfall is important in generation of Hydro Electric Power. The volume of water in rivers increases during the rainy season and this boost generation of electricity.
Morezactionszshouldzbeztakenztozsensitizezthezpubliczzaboutzzthezoccurrences ofwweatherzevents,zwhichziszfrequentznowadayszaszevidencezzofzzchangezzinzzthe climate (Ayanlade et al., 2020). Rainfall in Nigeria is produced by the intertropical discontinuity. Tropical analysts have consequently identified the boundary by several names such as The Intertropical Convergence Zone (ITCZ), Intertropical Fronts (ITF), and recently the Intertropical Discontinuity (ITD). The ITD is a warm, moist maritime air masses yielding heavy rainfall. Rainfall is copious in all mouths with an annual total often being 250 cm (Strahler & Strahler, 1988). It is primarily a region of "maximum" Surface moisture gradient, known as a humidity discontinuity (Oyewo, 2005).
Reductionzinzvolumezofzrainfallzandzsignificantzincreasezinzsurfaceztemperaturezwith thezfarmerszhavingzfirmzperceptionzofzthesezchanges (Tarfa et al., 2020). Warmer temperatures are very likely to produce more vigorous variability in climate such as increase evaporation, capacity of air to hold more moisture and thus heavier rainstorms.
Growing industrialization and increasing use of fossil fuels are putting pressure and affecting the regional and global temperatures that are subsequently influencing the overall precipitation levels. Increasingztemperaturezandzchangingzpatternszof precipitationzarezamongzthezmanyzconsequences,zwhichzzarezzattributedzztozzclimate change (Dammo et al., 2016). A change in temperature is an important indicator of global warming that directly determines the impact of climate change. Recent concern about rising global temperature was justified by its negative impact in all sectors of the economy most especially water supply, ecosystems, coastal habitats, and industries.
Furthermore, the report on cessation of the rainy season in the country indicated changes from "normal" between 1941 and 1970 to "early cessation" during the 1971 to 2000 period in most stations (NIMET, 2008). NIMET (2012) as certained that the period of the rainy season in the country has decline from 1941 while the signals of late onset and early cessation of the rainy season set in. Thus, the length of the rainy season has remained shrinking, the annual total rainfall is almost the same, thereby giving rise to occasional flash floods and drought occurrences during growing period. Temperatures across the country showed an increasing trend from mid-20 th century to date. The mean temperature anomaly indicated warming in most locations in the country. Temperatures have increased from 0.2 to 0.5 o C in the high ground areas of Jos, Yelwa and Ilorin in the north and Shaki, Iseyin and Ondo in the southwest to 0.9 to 1.9 o C over the rest parts of the country. Aiyelokun & Odekoya (2016)  Thezhighzvariationszofzrainfallzatzhigherzlatitudeszrevealzthezunreliableznaturezofzrain fallzaszonezprogressesztowardszNorthern Guineazandzvicezversa (Buba et al., 2017).
Thezobservedzspatiotemporalztrendszandzvariabilityzinzrainfallzarezimportantzbasiszfor guidingztargetingzofzappropriatezadaptivezmeasureszacrosszmultiplezsectors (Muthoni et al., 2019). There have been limited studies concerned on spatio-statistical analysis of rainfall and temperature. Therefore, this research intends to rainfall and temperature analysis based on onset, cessation and length of rainy season. This study aims to examine the variability of rainfall and temperature based on spatio-statistical analysis.

Study Area
The study area is Ijebu Ode in ogun state, Nigeria (Figure 1 and 2). The area experiences humid tropical climate which is characterized by alternate wet and dry season seasons like the rest of Nigeria. Ijebu-Ode region on annual basis is under the influence of hot-wet tropical maritime air mass during the rainy season (April-October) and hot-dry tropical continental air mass during the dry season (November-March) (Aiyelokun & Odekoya, 2016). Rainfall is generally heavy with peaks occurring in July and September (double maxima) coupled with high temperature, high evapotranspiration and high relative humidity. The average monthly rainfall for the area ranges between 7.1mm in the month of January to 208.27 mm in the month of June. The annual rainfall is between 1575 mm and 2340 mm. The temperature of the area ranges from 23 o C during the dry season to 35 o C during the rainy season with an average annual temperature is 27.5 o C. Furthermore, the area experiences relative humidity of 63 % in the dry season to as high as 95 % during the peak of the rainy season (Onakomaiya, 2000). industries producing high quality beer, bicycle tyres, ceramic goods, high quality clay bricks, carpet and clothing materials (Onakomaiya, 2000). The data required for the study include Mean and annual temperature, rainfall and sunshine hours between 1983-2017.
These data were collected from the Nigeria Meteorological Agency (NIMET) at the headquarters in Oshodi Lagos state while the map of Ijebu ode was extracted from the administrative map of Nigeria using Arc GIS 10.3 software. Onset of the rainy season will be compiled using Walter (1967) formula because of its higher reliability in predicting the onset of the rains among different methods. The formula is expressed in TM Where DM = number of days of the month containing the onset of rainfall, A = Total rainfall for the previous month, TM = total rainfall for the month in which 51 mm or more is reached.
The cessation of rainy season is defined as the last occasion of rainfall that record rainfall of 51 mm and above. Cessation of the rainy season will be compiled using Walter (1967) formula because of its higher reliability in predicting the onset of the rains among different methods. The formula is expressed in equation 2.
TM Where DM = number of days of the month containing the cessation of rainfall, A = Total rainfall for the previous month, TM = total rainfall for the month in which 51mm or more is reached. Therefore, the length of the rainy season is the total number of days between the Onset and Cessation date.
The Anomaly Index expresses the degree of rainfall and temperature anomaly for the relevant periods in relation to the long term mean rainfall and temperature for the study period. To calculate the rainfall and temperature anomaly for the study period, standardize rainfall anomaly index (SAI) will be used. The formula is given in equation Where x = mean annual rainfall or temperature, X = mean of entire series, SD = Standard deviation of the entire series. Correlation will be used to measure the relationship between the rainfall and temperature for the study period.
T test was used to ascertain whether there is a significant difference between onset dates, cessation dates and length of rainy season in the area. Time series is defined as a series of observation assumed by a variable over successive time periods. Time series analysis helps to fit an array of time bound data on a line of best fit, it helps to show the type of trends existing in the data graphically. Time series will be used to examine the trends of rainfall and temperature for the study period and for predict for the future years. The trend line equation of a time series data is as shown in equation 5.

Results and Discussion
In  Cessation dates in 1982Cessation dates in , 1983Cessation dates in , 1985Cessation dates in , 1986Cessation dates in , 1987Cessation dates in , 1989Cessation dates in , 1991Cessation dates in , 1993Cessation dates in , 1995Cessation dates in , 1996Cessation dates in , 2000Cessation dates in , 2001Cessation dates in , 2003Cessation dates in , 2004Cessation dates in , 2005Cessation dates in , 2006 and 2008 are lower than the mean value of 23 rd October while cessation dates in 1981, 1984, 1988, 1990, 1997, 1999. 2007, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016 and 2017 are higher than the mean value and cessation dates in 1992, 1994 and 1998 are the same dates with the mean value of 23 rd October. As shown in Table 1, the lowest length of rainy season of 159 days was recorded in 1983 while the highest of 261 days was recorded in 1997. The mean length of rainy season in the area is 223 days with a standard deviation of 20.8 and a coefficient of variation of 9.3%. This implies that the length of rainy season in the area is homogeneous. Furthermore, length of rainy season in 1983, 1985, 1992, 1994, 1998, 2001, 2002, 2005, 2007, 2008, 2010 and 2017 are lower than the mean value of 223 days while length of rainy season in 1981, 1982, 1984, 1986, 1987, 1988, 1989, 1990, 1991, 1993, 1995, 1996, 1997, 1999, 2000, 2003, 2004, 2006, 2009. 2011, 2012, 2013, 2014, 2015 and 2016 are higher than the mean value of 223 days. Figure 3 presents the pattern of Onset date of rainy season in the area during the study period. The trend of onset date is upward sloping, this implies that rainfall is starting late in the study period. The trend also shows that the latest date of onset was in 1983 while the earliest date was in 2004. This result implies that onset dates in the area is affected by climate change, rain now starts late in the study area.

Pattern of Cessation Dates
As shown in Figure 4, the trend of cessation dates in the area is upward sloping, this implies that cessation dates in the study area is getting late. Furthermore, the earliest cessation date was in 1983 while the latest was in 1990. Also, the trend shows that cessation dates where early between 1981 to 1996 and late between 1997 and 2017. This result shows that climate change has led to late cessation of rainfall in the area.  In Figure 6 , 1982, 1983, 1984, 1986, 1990, 1991, 1992, 1993, 1994, 1995, 1998, 2001, 2003, 2004, 2005, 2006 and 2017 experienced negative anomaly, this implies that these years are dry experiencing rainfall below the normal while 1981,1985,1987,1988,1989,1996,1997,1999,2000,2002,2007,2008,2009,2010,2012,2013,2014,2015 and 2016 experienced positive anomaly, rainfall received in this years are above the normal. Figure 6. Anomalous rainfall Figure 7 presents the graph of temperature anomaly in the area for the study period. Temperature recorded in 1981Temperature recorded in , 1982Temperature recorded in , 1983Temperature recorded in , 1985Temperature recorded in , 1986Temperature recorded in , 1989Temperature recorded in , 1992Temperature recorded in , 2005Temperature recorded in , 2007Temperature recorded in , 2008Temperature recorded in , 2009Temperature recorded in , 2010Temperature recorded in and 2011Temperature recorded in showed a negative anomaly while 1987Temperature recorded in , 1990Temperature recorded in , 1993Temperature recorded in , 1995Temperature recorded in , 1996Temperature recorded in , 1997Temperature recorded in , 1998Temperature recorded in , 2000Temperature recorded in , 2001Temperature recorded in , 2002Temperature recorded in , 2003Temperature recorded in , 2004Temperature recorded in , 2012Temperature recorded in , 2014Temperature recorded in , 2015 and 2016 showed a negative anomaly. Also, 1984Also, , 1988Also, , 1994Also, , 1999 and 2013 experienced a normal temperature for the study area. and Onset (r = -0.119) are negative; this implies that maximum temperature, average temperature and onset are inversely related to rainfall amount. If maximum temperature, average temperature and onset increases, rainfall decreases and if maximum temperature, average temperature and onset decreases, rainfall increases. The implication of this maximum temperature and average temperature influences rainfall in the area. Table 3, the correlation coefficient of maximum temperature (0.325), sunshine hours (0.380), average temperature (0.867), cessation and length of rainy season (0.080) is positive. This means that there is a directly relationship between minimum temperature and maximum temperature, sunshine hours' average temperature and length of rainy season. However, implies that if these variables increase, minimum temperature will increase and if these variables decrease, minimum temperature will decrease. Also, Onset (-0.118) and cessation (-0.026) is negative; this implies that there is an inverse relationship between onset date, cessation date and minimum temperature. Furthermore, the correlation coefficient Average temperature (r = 0.749) and

As indicated in
Length of rainy season (r = 0.131) is positive, this implies that there is a direct relationship between average temperature, length of rainy season and maximum temperature. Also, sunshine hours (r = -0.389), onset (r = -0.203) and cessation (r = -0.031) is negative, this implies that there is an inverse relationship between sunshine hours, onset, cessation and maximum temperature.
In the same vein, the correlation coefficient average temperature ( and average temperature (0.867*) is significantly related to minimum temperature. In the same vein, sunshine hours (-0.389*) and average temperature (0.749*) are significantly related to maximum temperature. Also, Onset (-0.642*) and cessation (0.509*) are significantly related to length of rainy season.  ThesezresultszarezsupportedzbyzMeshramzetzal.z(2018)zthatzthezzmonsoonzzandzzthez winter seasonzexhibited aznegativeztrendzinzrainfallzchangeszoverzthezperiodzofzstudy.

Conclusion
The area is experiencing a significant increase in minimum temperature, sunshine hours, rainfall and average temperature while it is experiencing a significant decrease in maximum temperature. However, rains start early in the area and ceases late, resulting to a prolonged length of rainy season because the area is experiencing a decline in the length of rainy season. It could also be noted that onset dates of rains and cessation dates are major determinants of length of rainy season as a change in any of the two variables will affect the length of the rainy season. The rainfall anomaly over all the area revealed that there was a composite nature in which. Early warning systems about extreme temperature and rainfall events should be put in place. This will help reduce the cost of destruction caused by this extreme weather events especially in the case of flood. Early warning about temperature events can help prevent disease such as meningitis that is related with temperature. The response should involve flood forecasting and early warning using rainfall data, rescue and evacuation and post flood impact assessment, recovery and rehabilitation.