J. Nig. Soc. Phys. Sci. 4 (2022) 820 Journal of the Nigerian Society of Physical Sciences Modeling and Forecasting Selected Meteorological Parameters for the Environmental Awareness in Sub-Sahel West Africa Stations F. O. Awedaa,∗, J. A. Akinpelua, T. K. Samsonb, M. Sannic, B. S. Olatinwoa aPhysics and Solar Energy Programme, College of Agriculture, Engineering and Science, Bowen University, Iwo, Osun State, Nigeria bStatistics Programme, College of Agriculture, Engineering and Science, Bowen University, Iwo, Osun State, Nigeria cDepartment of Science Technology, Federal Polytechnic, Offa, Kwara State Abstract The monthly air temperature, rainfall, air pressure, and wind speed direction for the environmental time series recorded between January 1, 1980 and December 31, 2020 in six African stations from different climatic zones were modeled and forecasted. In the forecasting, Augmented Dickey Fuller test, ARIMA models, Auto correlation Function (ACF) and Partial Autocorrelation Function (PACF) were used. Result showed that in most of the fitted models, the Moving Average terms both seasonal and non- seasonal were also significant (p < 0.05) indicating that the previous day value of the stochastic term also had a significant effect on the present value of meteorological parameters in the environment. It was observed that in all the fitted models except for wind direction in Conakry and rainfall in Abidjan have all their Autoregressive term of order 1 significant (p < 0.05) which implies that previous day value of these meteorological parameter had a significant effect on the present day value of the parameters. Therefore, the forecast model indicates that maximum temperature are expected in February, March, April, and June while minimum temperatures in January, August, December. Although, the selected models cannot forecast the precise air temperature, this can also provide information that can be of help to create tactics for appropriate preparation of farming which can be used as tools for effective environmental preparation and policymaking. DOI:10.46481/jnsps.2022.820 Keywords: Meteorological Parameters, ARIMA models, ACF, PACF, Forecast Article History: Received: 18 May 2022 Received in revised form: 06 July 2022 Accepted for publication: 25 July 2022 Published: 12 August 2022 c© 2022 The Author(s). Published by the Nigerian Society of Physical Sciences under the terms of the Creative Commons Attribution 4.0 International license (https://creativecommons.org/licenses/by/4.0). Further distribution of this work must maintain attribution to the author(s) and the published article’s title, journal citation, and DOI. Communicated by: O. J. Abimbola 1. Introduction The forecasting of future weather sequence variables for en- vironmental awareness is based on the spell of historical catego- rization, which is known as an essential metric for atmospheric ∗Corresponding author tel. no: +234(0)8032566053 Email address: aweda.francis@bowen.edu.ng (F. O. Aweda ) parameter forecasting and modeling [1-9]. Furthermore, agri- cultural products are subject to meteorological, ecological, and environmental variation [8, 10-14], wherein planetary and chrono- logical qualities have a very important link over the environ- ment by applying time series modeling i.e ARIMA model [5, 15-18]. International heating impacts, according to [19-22], contribute to the development of the atmospheric structure; nev- ertheless, this effect is on the upcoming alterations in the cli- 1 Aweda et al. / J. Nig. Soc. Phys. Sci. 4 (2022) 820 2 matic growth of the environment, where the occurrence and size of diverse happenings affect the growth of any region [6, 23]. According to the international heating effect, an increase in temperature and a decrease in rainfall lead to famine in any en- vironment; nevertheless, this contributes to multiple pressures on food as a result of reduced rainfall and an increase in envi- ronmental temperature [17]. The impact of weather quantity forecasting on environmen- tal consciousness is critical for agricultural products employ- ing statistical models for forecasting [4, 6, 24, 25]. Statistical time series offered historical information on the modeling of the atmospheric awareness of the atmospheric parameter, ac- cording to several studies [6]. Various writers have noticed that data series patterns play a role in forecasting based on the envi- ronment, which has an impact on the climatological influence of environmental degradation and development [4, 25]. The autoregressive integrated moving average (ARIMA) time se- ries modeling has been found to be a good method for fore- casting and predicting some atmospheric characteristics that is widely accepted [7, 24]. Furthermore, ARIMA models have been shown to be one of the most widely utilized models for a variety of applications, including science, technology, econ- omy, market, commerce, and industry[4, 6, 7]. ARIMA mod- els, in particular, have been found to be effective in forecasting various weather conditions [26]. [20] revealed linear ARIMA model and quadratic model as the overall best performing model for the prediction of the monthly and annual temperature in Libya. Following [27], the use of the ARIMA model for 50- years (1955 - 2005) in Shiraz, south of Iran was found to be a good model in the future forecasting of temperature. Moreover, [27, 28] use another model Kwon as a seasonal autoregres- sive integrated moving average (SERIMA) to forecast the mean monthly average of the maximum temperature experienced in the India sub-region. According to the above researchers, their results revealed that maximum temperature forecast was a good parameter for guiding agricultural producers. In another vein, [29] used different ARIMA models and linear trends models to forecast the temperature and precipitation in the Afyonkarahisar area in Turkey and found that there would be an increase in tem- perature till the year 2025. According to pieces of literatures, [30] using statistical properties studied the historical effect of temperature in Canada for the period 1913-2013 whereby deter- mine a seasonal ARIMA model to forecast forthcoming temper- ature records. In their research [31] implemented the use of the SERIMA model as proposed by different researchers on the fre- quency analysis and forecasting monthly rainfall in Umuahia. [7, 32, 33] predicted once-a-month rainwater across several ar- eas within Ghana using different models such as SERIMA and ARIMA models. Weekly and monthly rainfall using time series with the help of SERIMA models over selected weather stations in Malaysia were erected by [34] and in India according to [34, 35]. Various authors, including ([25, 36-40]), worked on atmo- spheric and meteorological modeling using various statistical models. The stated ARIMA versions take a good quality post- sample predicting performing for yearly and monthly agrome- teorological time series [6]. However, for this study trend pa- rameters shall be fitted for the polynomial function, and the pe- riod constraint will be projected along with Fourier sequences for the environmental awareness. Furthermore, the major pur- pose of this research was to examine statistical modelling of the monthly atmospheric parameters’ perception utilizing time se- ries for the environmental awareness from six distinct stations in Sub-Sahel African cities over four decades. 2. Methodology 2.1. Study Area This study was conducted for locations in the West African sub-region as shown in Figure 1. The selected locations for this research were: Daka (17.366 oW, 14.765 o N), Conakry (13.578 oW, 09.641 o N), Abijan (04.008 oW, 05.360 o N), Ba- mako (08.003 oW, 12.639 o N), Niamey (12.125 oW, 13.512 o N), Abuja (07.399 oW, 09.077 o N). This study used the data from 1980 to 2020 for all the stations considered. 2.2. Data Collection and Analysis The data for this study was collected from the archive of the Modern-Era Retrospective analysis for Research and Ap- plications, Version 2 (MERRA-2) Web services. The data were accessed on 5th of August, 2021 and the method of data col- lection followed what was done by [41] as reported by [24, 42, 43]. 2.3. Statistical Analysis Modeling The stationarity of each of these meteorological parameters in each station was tested using the Augmented Dickey Fuller test and p-value less than 0.05 indicates stationarity. The tenta- tive ARIMA models were identified based on Autocorrelation Function (ACF) and Partial Autocorrelation Function (PACF) plots and the estimation of parameters of the ARIMA were fa- cilitated using the Statistical Package for Social Sciences (SPSS version 20.0). After the estimation of the model parameters, model diagnostic checking principally aimed at checking the fitness of the models was determined based on Ljung-Box test with p-value greater than 0.05 indicating a good fit and hence one year forecast of these meteorological parameters were pro- vided for the different locations. 3. Results and Discussion 3.1. The variation of the meteorological parameters over the environment The results of the monthly variation of air temperature, rel- ative humidity, pressure, wind speed, wind direction and rain- fall for the African stations considered as related to tempera- ture are shown in Figures 2, 3, 4, 5, 6, and 7. Abidjan Figure (Aa, Ba, Ca, Da, Ea.) revealed that air temperature had a sig- nificant effect on the other parameters considered, as shown in Figure 2(Aa), the temperature got to it minimum T = 24.4 oC in August, while relative humidity got to it minimum RH = 71.9% in January and maximum temperature T = 27.9 oC was recorded in February while the maximum relative humidity 2 Aweda et al. / J. Nig. Soc. Phys. Sci. 4 (2022) 820 3 Table 1. The division of the studied African stations into the hinterland and coastal regions. Station Division Country Longitude Latitude Period of Data (oW) (o N) Dakar Coastal Region Senegal 17.366 14.765 1980 - 2020 Conakry Coastal Region Guinea 13.578 09.641 1980 - 2020 Abidjan Coastal Region Cote d’Ivoire 04.008 05.360 1980 - 2020 Bamako Hinterland Region Mali 08.003 12.639 1980 - 2020 Niamey Hinterland Region Niger 12.125 13.512 1980 - 2020 Abuja Hinterland Region Nigeria 07.399 09.077 1980 - 2020 Figure 1. Map of Africa showing the stations for the research. RH = 85.2% was recorded in June. Figure 2(Ba) shows the cor- relation between temperature and pressure. The results revealed that the maximum temperature T = 27.9 oC was recorded in February while the maximum air pressure P = 1009.9 hPa has been recorded in July. However, the minimum tempera- ture was recorded in August T = 24.4 oC and the minimum air pressure was recorded in February P = 1006.0 hPa. How- ever, the maximum and minimum temperature of the station does not change for the plot across all other parameters. Figure 1(Ca) showed that wind speed had a minimum value in Jan- uary (WS = 2.2 m/s), while the maximum value was recorded in July (WS = 3.9 m/s). This correlation shows that, when the temperature is low, there is an increase in the wind speed of the station. Figure 2(Da) shows that the wind direction got to its minimum in November (W D = 189.1o), this indicates that the direction towards which the wind is moving was low in November, maximum in September (W D = 209.0o). Max- imum rainfall was recorded in May (RF = 253.8 mm) while the minimum rainfall was recorded in August (RF = 84.0 mm). The result demonstrates that temperature has a significant influ- ence on the other parameters, as the higher the temperature, the greater the variation in the other parameters. This applies to all of the stations in this investigation. Figure 3 (Ab, Bb, Cb, Db, Eb) shows the correlation be- tween temperature and other parameters. The result Ab shows that maximum temperature T = 27.7 oC was observed in March, while minimum temperature T = 23.8 oC was observed in Au- gust. This shows that August is more humid than other months. However, Maximum relative humidity RH = 85.8% was ob- served in August while minimum relative humidity RH = 43.6% was observed in February. The correlation between the two pa- rameters shows that when the temperature is at maximum, rel- ative humidity at minimum, vice vasa. Figure 3(Bb) shows the relationship between temperature and air pressure, this revealed 3 Aweda et al. / J. Nig. Soc. Phys. Sci. 4 (2022) 820 4 Figure 2. The Environmental Variation of Atmospheric Parameters for Abidjan. Table 2. Summary result for Augmented Dickey Fuller (ADF) Test for stationarity of the meteorological parameters. Station Pressure Rainfall RH Temp WD WS Dakar 0.0054* 0.0013* 0.0009* 0.0020* 0.0006* 0.0001* Conakry 0.0011* 0.0036* 0.0023* 0.0000** 0.0000* 0.0104* Abidjan 0.002* 0.0023* 0.0015* 0.0048* 0.0227* 0.0082* Bamako 0.0104* 0.0044* 0.0000** 0.0000** 0.0159* 0.0001* Niamey 0.0003* 0.0051* 0.003* 0.0000* 0.0000* 0.0001* Abuja 0.0007* 0.0010* 0.0088* 0.0213 0.0001* 0.0001* RH- Relative humidity, WD- Wind direction, WS- Wind speed, values reported are the p-values, **stationary after first differencing, *stationary at level that as the temperature increases air pressure decreases. How- ever, air pressure (P = 961.4 hPa) got to its minimum in March and maximum (P = 964.7 hPa), this shows that air pressure is more pronounced in March at Abuja than in other months. The correlation of wind speed and temperature as shown in Figure 3(Cb), revealed that wind speed was at minimum level in Oc- 4 Aweda et al. / J. Nig. Soc. Phys. Sci. 4 (2022) 820 5 Figure 3. The Environmental Variation of Atmospheric Parameters for Abuja. tober and maximum in April as shown. More so, for wind di- rection, the maximum value was recorded in August and the minimum in January. This show that the wind direction is high in August than in other months as revealed by Figure 3(Db). For Figure 3(Eb), the rainfall got to its maximum (RF = 556.5 mm) in August with a sharp increase, whereas there was no record of rainfall or zero rainfall in January, November, and Decem- ber. This was also reported by [31]. These months signify the winter period of the station, where no record of rainfall was ob- served. The results of Abuja revealed that temperature follows a sinusoidal pattern with a minimum in August and a maximum in March. This shows the transition between the winter to the rainy season of the station. The results of the station Bamako as presented in Figures Ac, Bc, Cc, Dc, Ec and Fc, revealed that temperature has a min- imum and maximum values as T = 32.5 oC and T = 23.4 oC 5 Aweda et al. / J. Nig. Soc. Phys. Sci. 4 (2022) 820 6 Figure 4. The Environmental Variation of Atmospheric Parameters for Bamako. in January and April respectively. This signifies the winter and the summer month for the station. The result also shows that relative humidity has minimum and maximum values as RH = 18.3 % in February and RH = 81.5 % in August, this shows that the water content in the atmosphere is low in Febru- ary and high in August. Thereby, August is classified to be the rainy season for the station. The air pressure was observed to be in the range (965.5 - 968.4) hPa, this shows that the air pressure was observed to be significant in the station. Wind speed was observed to be in the range (0.6 and 3.9) m/s, this shows that 6 Aweda et al. / J. Nig. Soc. Phys. Sci. 4 (2022) 820 7 Figure 5. The Environmental Variation of Atmospheric Parameters for Conakry. the wind speed for the station is low to power any wind turbine machine in the station. The direction of the wind shows that the wind moves with the value (54.0 - 240.8) o. Rainfall is at maximum of 227 mm in August and a minimum of 0.6 mm in December. This shows that the rate of rainfall in August is as high as the other months. August is observed to be the peak of rainfall in the station, and it signifies the summer month. The station Conakry shows that temperature ranges between 7 Aweda et al. / J. Nig. Soc. Phys. Sci. 4 (2022) 820 8 Figure 6. The Environmental Variation of Atmospheric Parameters for Dakar. T = (25.7 − 27.4) oC, the relative humidity is at its minimum 62.6 % in January and maximum in August 86.5 %, this shows that the water content in the atmosphere is more in August and low in January. The air pressure of the station showed that June 1009.5 hPa had the maximum value as compared to the other months, whereby, March had the minimum 1006.5 hPa. The wind speed and wind direction revealed that December had the minimum values with May and August as the highest. The rain- fall of 798.0 mm in July was the highest and 6.6 mm was the lowest in the station. 8 Aweda et al. / J. Nig. Soc. Phys. Sci. 4 (2022) 820 9 Figure 7. The Environmental Variation of Atmospheric Parameters for Niamey. Figure 6(Ae, Be, Ce, De, Ee.) revealed that air temper- ature has a significant effect on other parameters considered, as shown in Figure 6(Ae), the temperature got to it minimum T = 20.0 oC in January, while relative humidity got to it min- imum RH = 12.4% value in April and maximum temperature T = 34.3 oC was recorded in June while maximum relative humidity RH = 41.6% was recorded in September. Figure 6(Be) shows the correlation between temperature and pressure. The results revealed maximum air pressure P = 980.5 hPa was recorded in January and minimum air pressure was recorded in April P = 974.2 hPa. However, the maximum and minimum temperature of the station does not change for the plot across all other parameters. Figure 6(Ca) shows that wind speed has a minimum value in June (WS = 12.2 m/s), while the maximum value was recorded in February (WS = 5.2 m/s). This corre- lation shows that, when the temperature, there is an increase in 9 Aweda et al. / J. Nig. Soc. Phys. Sci. 4 (2022) 820 10 Table 3. Summary result for ARIMA models for the different meteorological parameters in the various stations. Stations Models Constant AR1 AR2 MA1 MMA2 MA3 SAR1 SMA1 Ljung Box test p-value Temp ARI MA(1,0,1)(1,0,1)12 28.091 0.677 0.406 - - 1.000 0.903 0.288 (0.000)** (0.000)** (0.000)** (0.000)** (0.000)** RH ARI MA(1,0,0)(1,0,1)12 25.893 0.622 - - - 0.998 0.874 0.507 (0.000)** (0.000)** (0.000)** (0.000)** P ARI MA(1,0,1)(1,0,1)12 977.164 0.274 - - - - 0.999 0.921 0.195 Dakar (0.000)** (0.000)** (0.000)** (0.000)** WS ARI MA(1,0,1)(1,0,1)12 3.610 1.115 - - - - 0.999 0.920 0.492 (0.000)** (0.011)* (0.000)** (0.000)** WD ARI MA(0,0,0)(1,0,1)12 92.466 - - - - 0.999 0.898 0.127 (0.000)** - - - - (0.000)** (0.000)** RF ARI MA(1,0,0)(1,0,1)12 6.140 0.238 0.992 0.857 0.908 (0.103) (0.000)** (0.000)** (0.000)** Temp ARI MA(1,1,1)(1,0,1)12 0.001 0.599 0.9999 - - 0.9999 0.935 0.691 (0.000)** (0.000)** (0.000)** (0.000)** (0.000)** RH ARI MA(1,0,1)(1,0,1)12 1007.848 0.838 0.563 - - 0.999 0.926 0.227 (0.000)** (0.000)** (0.000)** (0.000)** (0.000)** P ARI MA(1,0,1)(1,0,1)12 -0.005 0.365 0.053 - - -0.439 0.990 0.634 Conakry (0.651)** (0.008)** (0.718)** (0.000)** (0.004)** WS ARI MA(1,0,1)(1,0,1)12 2.533 0.204 - - - 0.998 0.845 0.053 (0.000)** (0.000)** (0.000)** (0.000)** WD ARI MA(1,0,1)(1,2,1)12 -0.017 -0.229 -0.068 - - -0.477 0.990 0.634 (0.784) (0.432) (0.818) (0.000)** (0.000)** RF ARI MA(1,0,0)(1,0,1)12 289.068 0.119 - - - 0.999 0.960 0.430 (0.000)** (0.009)** (0.000)** (0.001)** Temp ARI MA(1,0,1)(1,0,1)12 26.442 0.620 - 0.023 - - 0.999 0.895 0.432 (0.000)** (0.000)** (0.765) (0.000)** (0.000)** RH ARI MA(1,0,1)(1,0,1)12 1007.461 0.802 - 0.515 - - 0.999 0.920 0.065 (0.000)** (0.000)** (0.000)** (0.000)** (0.000)** P ARI MA(1,0,1)(1,0,1)12 80.66 0.516 - 0.044 - - 0.995 0.883 0.265 Abidjan (0.000)** (0.000)** (0.643) (0.000)** (0.000)** WS ARI MA(1,0,1)(1,0,1)12 3.060 0.390 - 0.100 - - 0.997 0.885 0.750 (0.000)** (0.006)** (0.508) (0.000)** (0.000)** WD ARI MA(1,0,1)(1,0,1)12 203.989 0.978 - 0.918 - - 0.999 0.967 0.145 (0.000)** (0.000)** (0.508) (0.000)** (0.000)** RF ARI MA(1,0,0)(1,0,1)12 154.539 0.264 - 0.033 - - 0.982 0.832 0.145 (0.000)** (0.1620) (0.867) (0.000)** (0.000)** Temp ARI MA(1,1,1)(1,1,1)12 -0.000286 0.2910 - 0.924 - - 0.023 0.998 0.143 (0.467) (0.000)** (0.000)** (0.662) (0.158) RH ARI MA(1,1,1)(1,1,1)12 0.004 0.5290 - 0.955 - - 0.139 0.996 0.894 (0.467) (0.000)** (0.000)** (0.008)** (0.000)** P ARI MA(1,0,1)(1,0,1)12 967.146 0.7430 - 0.440 - - 0.999 0.939 0.245 Bamako (0.000)** (0.000)** (0.000)** (0.008)** (0.000)** WS ARI MA(2,0,3)(1,0,1)12 2.129 1.711 -0.951 1.548 -0.711 -0.084 0.999 0.917 0.052 (0.000)** (0.000)** (0.000)** (0.000)** (0.000)** (0.087) (0.008)** (0.000)** WD ARI MA(1,0,0)(1,0,1)12 141.407 0.104 - - - - 0.995 0.766 0.131 (0.000)** (0.022)** (0.000)** (0.000)** RF ARI MA(1,0,0)(1,0,1)12 56.648 0.148 - - - - 0.999 0.969 0.131 (0.012)** (0.001)** (0.000)** (0.000)** Temp ARI MA(1,0,1)(1,0,1)12 28.674 0.555 0.254 - - 0.999 0.903 0.511 (0.000)** (0.000)** (0.048)* (0.000)** (0.000)** RH NSM P ARI MA(1,0,1)(1,0,1)12 982.790 0.627 0.353 - - 0.999 0.964 0.905 Niamey (0.000)** (0.000)** (0.004)** (0.000)** (0.000)** WS ARI MA(0,0,0)(1,0,1)12 - - - - - 0.999 0.9160 0.561 (0.000)** (0.000)** WD ARI MA(0,0,0)(1,0,1)12 141.247 0.034 - - - 0.999 0.983 0.300 (0.000)** (0.460)** (0.000)** (0.000)** RF ARI MA(1,0,0)(1,0,1)12 25.059 0.201 - - - 0.9930 0.813 0.568 (0.019)* (0.000)** (0.000)** (0.000)** Temp ARI MA(1,1,1)(1,0,1)12 25.090 0.810 0.217 - - 0.997 0.869 0.187 (0.000)** (0.000)** (0.001)** (0.000)** (0.000)** RH ARI MA(1,0,1)(1,0,1)12 68.759 0.710 0.182 - - 0.998 0.885 0.360 (0.000)** (0.016)* (0.000)** (0.000)** (0.000)** P ARI MA(1,0,1)(1,0,1)12 963.248 0.770 0.485 - - 0.998 0.900 0.387 Abuja (0.000)** (0.000)** (0.000)** (0.000)** (0.000)** WS ARI MA(1,0,1)(1,0,1)12 1.498 0.461 0.283 - - 0.999 0.972 0.750 (0.000)** (0.022)* (0.192) (0.000)** (0.000)** WD ARI MA(1,0,0)(1,2,1)12 169.613 -0.022 -0.178 - - 0.999 0.931 0.142 (0.000)** (0.000)* (0.000)** (0.000)** (0.000)** RF ARI MA(1,0,0)(1,0,1)12 158.924 - - - - 0.999 0.909 0.182 (0.000)** (0.000)** (0.000)** *Significant at 5% (p < 0.05), **Significant at 1% ( p < 0.01), NSM- No Suitable Model the wind speed of the station. Figure 6(Da) shows that the wind direction got to its minimum in February (W D = 31.8o), this indicates that the direction towards which the wind is moving was low in February. However, it got to its maximum in July 10 Aweda et al. / J. Nig. Soc. Phys. Sci. 4 (2022) 820 11 Table 4. Summary of forecast of meteorological parameters in 2021. Months January February March April May June July August September October November December Dakar 20.33 23.64 28.21 31.95 34.21 34.60 32.57 29.92 30.80 29.36 24.78 20.82 Conakry 28.23 28.13 27.96 27.99 27.91 27.23 26.45 26.12 26.46 27.11 27.73 27.68 Temperature Abidjan 27.31 27.86 27.71 27.51 27.06 26.01 24.94 24.63 25.30 26.16 26.92 27.16 Bamako 23.91 26.46 29.74 32.27 31.88 29.18 26.45 25.23 25.57 26.54 25.48 23.31 Niamey 23.42 26.54 30.50 33.79 34.23 32.11 29.52 27.62 28.07 29.18 26.96 23.93 Abuja 25.41 27.67 29.11 28.54 26.92 25.64 24.62 24.15 24.60 25.13 25.31 24.40 Dakar 19.33 15.11 12.45 11.72 16.38 26.09 40.40 55.84 45.09 28.65 20.63 22.02 Conakry 63.49 65.15 67.34 70.61 77.84 81.63 85.64 86.36 84.90 81.98 74.65 67.44 Relative Abidjan 75.69 77.61 80.70 82.84 84.22 85.06 84.80 83.46 83.65 83.51 82.04 77.92 Humidity Bamako 28.73 22.87 23.32 34.08 46.91 64.15 78.73 86.56 84.18 71.25 46.92 35.20 Niamey - - - - - - - - - - - - Abuja 41.16 40.71 51.24 64.31 76.49 81.21 83.92 85.26 83.98 79.43 61.78 46.55 Dakar 980.41 978.62 976.49 974.67 974.69 975.38 976.17 977.12 977.06 977.17 978.65 980.39 Conakry 1007.48 1006.82 1006.68 1006.75 1007.79 1009.06 1009.37 1009.15 1008.56 1007.75 1007.29 1007.61 Pressure Abidjan 1006.70 1006.01 1005.97 1006.12 1007.28 1008.89 1009.50 1009.38 1008.56 1007.47 1006.65 1006.77 Bamako 968.63 967.15 965.69 964.76 965.75 967.45 968.06 967.94 967.99 967.37 967.43 968.52 Niamey 985.04 983.42 981.33 979.79 980.66 982.48 983.30 983.55 983.46 982.72 983.30 984.76 Abuja 963.25 962.08 961.55 961.69 963.03 964.43 964.75 964.60 964.28 963.37 962.84 963.32 Dakar 4.98 5.15 4.89 4.04 2.29 1.40 3.42 3.00 1.53 3.01 4.70 5.06 Conakry 1.12 2.27 3.29 3.45 3.13 3.11 4.36 4.60 2.93 1.61 0.69 0.87 Wind Abidjan 2.13 2.98 3.17 2.96 2.89 3.50 3.80 3.87 3.59 3.07 2.36 1.86 Speed Bamako 3.75 3.61 2.68 1.02 1.79 2.01 1.64 1.47 0.66 0.71 2.27 3.61 Niamey 4.13 3.83 2.85 1.38 3.12 3.61 3.15 2.22 1.68 1.41 2.84 3.91 Abuja 1.56 1.11 1.24 1.90 1.73 1.68 1.91 1.84 1.09 0.82 1.28 1.75 Dakar 34.25 33.26 35.77 52.80 66.87 166.12 217.38 209.63 178.39 67.88 43.54 37.41 Conakry 172.18 245.84 250.73 253.05 260.33 243.58 242.35 230.74 249.20 228.12 224.98 288.50 Wind Abidjan 207.00 213.61 214.48 212.81 206.76 206.54 211.48 213.79 214.00 208.84 194.38 198.62 Direction Bamako 60.00 57.05 51.42 82.24 191.22 210.36 227.14 242.40 165.71 130.56 67.29 63.67 Niamey 54.18 53.37 53.18 165.19 213.17 214.98 217.19 217.75 206.62 162.16 73.92 59.08 Abuja 66.78 122.09 193.64 217.06 220.62 222.25 231.96 230.92 192.51 153.75 92.82 71.67 Dakar 0.26 0.34 0.47 0.78 3.02 5.61 19.43 49.03 12.15 1.95 0.35 0.38 Conakry 10.30 9.09 14.28 29.71 181.35 466.29 799.55 766.82 608.51 426.43 135.00 19.42 Rainfall Abidjan 71.46 111.98 149.54 147.04 200.59 225.24 127.42 98.04 189.13 247.20 201.00 123.49 Bamako 2.83 1.75 3.04 12.42 29.57 58.56 145.64 227.11 135.31 47.73 4.33 1.32 Niamey 0.75 5.04 3.41 4.09 15.44 23.05 48.94 142.67 57.58 22.88 0.96 0.92 Abuja 2.26 7.53 17.90 71.01 157.21 207.67 407.30 503.16 295.58 129.21 8.07 2.16 (W D = 219.9o). Maximum rainfall was recorded in August (RF = 36.4 mm) while zero rainfall was recorded in February (RF = 0.0 mm). The maximum rainfall indicates that there is a low temperature during the period, this signifies increases in rainfall rate due to the low temperature for the station. The result shows that temperature has a significant effect on other parameters because of the more the temperature the higher the variation of other parameters. Figure 7 (Af, Bf, Cf, Df, Ef) shows the correlation between temperature and other parameters. The result Af shows that maximum temperature T = 34.1 oC was observed in April, while minimum temperature T = 23.2 oC was observed in Jan- uary. Maximum relative humidity RH = 70.3% was observed in August while minimum relative humidity RH = 18.0% was observed in January. The correlation between the two param- eters shows that when the temperature is at maximum, relative humidity at minimum, vice vasa. Figure 7(Bf) shows the re- lationship between temperature and air pressure, this revealed that as the temperature increases air pressure increases. How- ever, air pressure (P = 979.7 hPa) got to its minimum in Febru- ary and maximum (P = 984.9 hPa), this shows that air pres- sure is more pronounced in January at Niamey than in other months. As shown in Figure 7(Cf), the correlation of wind speed and temperature, revealed that wind speed was at min- imum in April and maximum in January. More so, for wind direction, the maximum value recorded was in August and the minimum in March. This show that the wind direction was high in August than in other months as revealed by Figure 7(Df). For Figure 3(Ef), the rainfall got to its maximum (RF = 93.1 mm) in August with a sharp increase, whereas there was no record of rainfall or zero rainfall in December. The results of Niamey revealed that temperature follows a sinusoidal pattern with the minimum in January and a maximum in April. This shows the transition between the dry to rainy seasons of the station. 3.2. Statistical Analysis Table 2 shows summary result of the stationarity of the se- ries using Augmented Dickey Fuller test for each of the meteo- rological parameter in the different locations. The result reveals that in all stations, all the parameters were stationary at level with exception of temperature in Conakry, RH and temperature in Bamako which were only stationary after first differencing meaning that they are integrated of order 1. The estimates of the different Autoregressive Integrated Mov- ing Average (ARIMA) models fitted to these meteorological pa- rameters are as shown in Table 3. The Ljung Box test was ap- plied to diagnose the different ARIMA model fitted and the p- values obtained for all models were greater than 0.05 (p > 0.05) which indicates that these models were of good fit. In all the fit- ted models except for wind direction in Conakry and rainfall in 11 Aweda et al. / J. Nig. Soc. Phys. Sci. 4 (2022) 820 12 Abidjan have all their Autoregressive term of order 1 signifi- cant ( p < 0.05) which implies that previous day value of these meteorological parameter has a significant effect on the present day value of the parameters. This indicates as the previous day value of these parameters increases significantly, there is a cor- responding significant increase in the present day value of the meteorological parameters. Result also shows that in most of the fitted models, the Moving Average terms both seasonal and non- seasonal were also significant (p < 0.05) indicating that the previous day value of the stochastic term also has a signifi- cant effect on the present value of meteorological parameters in the study area. Since the different ARIMA models were found to be of good fit, these models were used in forecasting the future values of the meteorological parameters in each of the selected locations and the summary result of the forecast are presented in Table 4. The forecast indicates the maximum temperature are ex- pected in June in Daka while in Conakry, Abidjan, Bamako, Niamey and Abuja, maximum temperature are expected in Jan- uary, February, April, May and March respectively. Also, min- imum temperatures were predicted in Daka, Conakry, Abidjan, Bamako, Niamey and Abuja in the months of January, August, December, January and August respectively. In Daka, Conakry, Bamako and Abuja, maximum relative humidity was predicted in August while in Abidjan maximum relative humidity was predicted in June. For other these meteorological parameters, the predicted values for 2021 were presented in Table 4. 4. Conclusion The environment and statistical analysis of the meteoro- logical parameters shows that the monthly mean variations of air temperature and other parameters considered for African stations exhibit similarity behaviour and dynamics, however, some statistical parameters fluctuate significantly between the stations studied. Air temperature, rainfall, air pressure, wind speed and direction modelling and its predicting posture a stim- ulating mission aimed at treatment slightly on monthly time series over the environment. In this study, it was observed that ARIMA model for the environmental study can proficiently get the reason of the increase in the meteorological parameter study by producing the minimum prediction of the root mean squared error which could be better forecast in the long time seasonal time series of the high frequency. For this research, the best model for prediction depends on the stations. The re- sults revealed different time series (ARIMA 101, ARIMA 111, ARIMA 110, ARIMA 011) which are appropriate for the West African continental data sets. However, investigative inspec- tion settles the competence of the models. Therefore, the fore- cast model indicates that maximum temperature are expected in June while minimum temperatures in January, August, Decem- ber. Although, the selected models cannot forecast the precise air temperature, this can also provide information that can be of help to create tactics for appropriate preparation of farming which can be used as tools for effective environmental prepara- tion and policymaking. Acknowledgments We thank the referees for the positive enlightening com- ments and suggestions, which have greatly helped us in making improvements to this paper. The authors will also like to ex- press their appreciation to HelioClim MERRA-2 for providing the data for analysis and Bowen University for the leave granted during the course of the research. References [1] F.O. Aweda & T.K. Samson, ”Modelling the Earth’s Solar Irradiance Across Some Selected Stations in Sub-Sahara Region of Africa”, Iranian (Iranica) Journal of Energy & Environment, 11 (2020) 204. [2] K. Lamorski, T. Pastuszka, J. Krzyszczak, C. Sławiński & B. Witkowska- Walczak, ”Soil water dynamic modeling using the physical and support vector machine methods”, Vadose Zone Journal, 12 (2013). [3] P. Baranowski, J. Krzyszczak, C. Slawinski, H. Hoffmann, J. Kozyra, A. Nieróbca, et al., ”Multifractal analysis of meteorological time series to assess climate impacts”, Climate Research, 65 (2015) 39. [4] M. Murat, I. Malinowska, H. Hoffmann & P. Baranowski, ”Statistical modelling of agrometeorological time series by exponential smoothing”, International Agrophysics, 30 (2016). [5] J. Krzyszczak, P. Baranowski, M. Zubik & H. Hoffmann, ”Temporal scale influence on multifractal properties of agro-meteorological time series”, Agricultural and Forest Meteorology, 239 (2017) 223. [6] M. Murat, I. Malinowska, M. Gos & J. Krzyszczak, ”Forecasting daily meteorological time series using ARIMA and regression models”, Inter- national agrophysics, 32 (2018). [7] A. Abdul-Aziz, M. Anokye, A. Kwame, L. Munyakazi & N. Nsowah- Nuamah, ”Modeling and forecasting rainfall pattern in Ghana as a sea- sonal ARIMA process: The case of Ashanti region”, (2013). [8] O. Abimbola, E. Utah & B. Alkali, ”Clmate Change. A Case Study of Jos, Nigeria”, African Journal of Physical Science, 4 (2011) 82. [9] R. EL CHAAL & M. Aboutafail, ”Statistical Modelling by Topological Maps of Kohonen for Classification of the Physicochemical Quality of Surface Waters of the Inaouen Watershed Under Matlab”, Journal of the Nigerian Society of Physical Sciences, (2022) 223. [10] S. Fronzek, N. Pirttioja, T. R. Carter, M. Bindi, H. Hoffmann, T. Palo- suo, et al., ”Classifying multi-model wheat yield impact response sur- faces showing sensitivity to temperature and precipitation change”, Agri- cultural systems, 159 (2018) 209. [11] N. Pirttioja, T. R. Carter, S. Fronzek, M. Bindi, H. Hoffmann, T. Palosuo, et al., ”Temperature and precipitation effects on wheat yield across a Eu- ropean transect: a crop model ensemble analysis using impact response surfaces”, Climate Research, 65 (2015) 87. [12] J. R. Porter & M. A. Semenov, ”Crop responses to climatic variation,” Philosophical Transactions of the Royal Society B: Biological Sciences, 360 (2005) 2021. [13] M. Ruiz-Ramos, R. Ferrise, A. Rodrı́guez, I. Lorite, M. Bindi, T. R. Carter, et al., ”Adaptation response surfaces for managing wheat under perturbed climate and CO2 in a Mediterranean environment”, Agricul- tural Systems, 159 (2018) 260. [14] A. Adedotun, T. Latunde & O. Odusanya, ”Modelling and Forecasting Climate Time Series with State-Space Model”, Journal of the Nigerian Society of Physical Sciences, (2020) 149. [15] H. Hoffmann, P. Baranowski, J. Krzyszczak, M. Zubik, C. Sławiński, T. Gaiser, et al., ”Temporal properties of spatially aggregated meteorological time series”, Agricultural and forest meteorology, 234 (2017) 247. [16] R. Walczak, B. Witkowska-Walczak, and P. Baranowski, ”Soil structure parameters in models of crop growth and yield prediction. Physical sub- models”, International agrophysics, 11 (1997). [17] D. B. Lobell, G. L. Hammer, G. McLean, C. Messina, M. J. Roberts & W. Schlenker, ”The critical role of extreme heat for maize production in the United States”, Nature climate change, 3 (2013) 497. [18] G. O. Odekina, A. F. Adedotun & O. F. Imaga, ”Modeling and Forecast- ing the Third wave of Covid-19 Incidence Rate in Nigeria Using Vec- tor Autoregressive Model Approach”, Journal of the Nigerian Society of Physical Sciences, (2022) 117. 12 Aweda et al. / J. Nig. Soc. Phys. Sci. 4 (2022) 820 13 [19] D. B. Lobell, A. Sibley & J. Ivan Ortiz-Monasterio, ”Extreme heat effects on wheat senescence in India”, Nature Climate Change, 2 (2012) 186. [20] E. El-Mallah & S. Elsharkawy, ”Time-series modeling and short term prediction of annual temperature trend on Coast Libya using the box- Jenkins ARIMA Model”, Advances res, 6 (2016) 1. [21] M. A. Semenov & P. R. Shewry, ”Modelling predicts that heat stress, not drought, will increase vulnerability of wheat in Europe”, Scientific Reports, 1 (2011) 1. [22] J. Sillmann & E. Roeckner, ”Indices for extreme events in projections of anthropogenic climate change”, Climatic Change, 86 (2008) 83. [23] S. O. Adams, D. A. Obaromi & A. A. Irinews, ”Goodness of Fit Test of an Autocorrelated Time Series Cubic Smoothing Spline Model”, Journal of the Nigerian Society of Physical Sciences, (2021) 191. [24] F. O. Aweda, S. Adebayo, T.K. Samson & I. A. Ojedokun, ”Modelling Net Radiative Measurement of Meteorological Parameters Using MERRA-2 Data in Sub-Sahara African Town”, Iranian (Iranica) Journal of Energy & Environment, 12 (2021) 173. [25] A. S. Gautam, D. Siingh & A. Kamra, ”Statistical analysis of the atmo- spheric ion concentrations and mobility distributions at a tropical station, Pune”, Quarterly Journal of the Royal Meteorological Society, 143 (2017) 2116. [26] F. O. Aweda, A. A. Adeniji, J. A. Akinpelu & A. J. Abiodun, ”Analysis of rainfall trends and variabilities for three decades in Sub–Sahara Africa”, Ruhuna Journal of Science, 12 (2021). [27] Y. BALYANI, G. FAZELNIYA & A. BAYAT, ”A study and prediction of annuel temperature in shiraz using ARIMA model”, (2012). [28] K. Anitha, N. Boiroju & P. Reddy, ”Forecasting of monthly mean of max- imum surface air temperature in India”, Int. J. Statistika Mathematika, 9 (2014) 14. [29] D. Afyonkarahisar & Y. V. S. Box, ”The Analyse Of Precipitation and Temperature in Afyonkarahisar (Turkey) in Respect of Box-Jenkins Tech- nique” [30] S. Khedhiri, ”Forecasting temperature records in PEI, Canada”, Letters in Spatial and Resource Sciences, 9 (2016) 43. [31] A. Akpanta, I. Okorie & N. Okoye, ”SARIMA modelling of the frequency of monthly rainfall in Umuahia, Abia state of Nigeria”, American journal of mathematics and statistics, 5 (2015) 82. [32] E. Afrifa-Yamoah, B. Saeed & A. Karim, ”Sarima modelling and fore- casting of monthly rainfall in the Brong Ahafo Region of Ghana”, World Environment, 6 (2016) 1. [33] F.O. Aweda, S. J. Olufemi & J. O. AGBOLADE, ”Meteorological Pa- rameters Study and Temperature Forecasting in Selected Stations in Sub- Sahara Africa using MERRA-2 Data”, Nigerian Journal of Technological Development, 19 (2022) 80. [34] F. Yusof & I. Kane, ”Modelling monthly rainfall time series using ETS state space and SARIMA models”, International Journal of Current Re- search, 4 (2012) 195. [35] P. Dabral & M. Z. Murry, ”Modelling and forecasting of rainfall time series using SARIMA”, Environmental Processes, 4 (2017) 399. [36] A. S. Gautam, K. Singh, M. Sharma, S. Gautam, A. Joshi & S. Kumar, ”Classification of Different Sky Conditions Based on Solar Radiation Ex- tinction and the Variability of Aerosol Optical Depth, Angstrom Expo- nent, Fine Particles Over Tehri Garhwal, Uttarakhand, India”, MAPAN, (2022) 1. [37] J. Thapliyal, M. Bhattacharyya, S. Prakash, B. Patni, S. Gautam, and A. S. Gautam, ”Addressing the relevance of COVID–19 pandemic in nature and human socio-economic fate”, Stochastic Environmental Research and Risk Assessment, (2022) 1, 2022. [38] J. V. Nepolian, D. Singh, R. Singh, A. S. Gautam & S. Gautam, ”Anal- ysis of Positive and Negative Atmospheric Air Ions During New Particle Formation (NPF) Events over Urban City of India”, Aerosol Science and Engineering, 5 (2021) 460. [39] B. Ambade, T. K. Sankar, A. Panicker, A. S. Gautam & S. Gautam, ”Char- acterization, seasonal variation, source apportionment and health risk as- sessment of black carbon over an urban region of East India”, Urban Cli- mate, 38 (2021) 100896. [40] S. Srivastava, A. Kumar, K. Bauddh, A. S. Gautam & S. Kumar, ”21-day lockdown in India dramatically reduced air pollution indices in Lucknow and New Delhi, India”, Bulletin of environmental contamination and tox- icology, 105 (2020) 9. [41] R. Gelaro, W. McCarty, M. J. Suárez, R. Todling, A. Molod, L. Takacs, et al., ”The modern-era retrospective analysis for research and applications, version 2 (MERRA-2)”, Journal of climate, 30 (2017) 5419. [42] F. O. Aweda, J. A. Oyewole, J. B. Fashae & T. K. Samson, ”Variation of the Earth’s Irradiance over Some Selected Towns in Nigeria”, Iranian (Iranica) Journal of Energy & Environment, 11 (2020) 301. [43] S. Adebayo, F. O. Aweda, I. A. Ojedokun & O. T. Olapade, ”Refractive Index Perception and Prediction of Radio wave through Recursive Neural Networks using Meteorological Data Parameters”, 35 (2022) 810. 13