1 Introduction

Rain-fed agriculture is the main mode of agronomic production undertaken by smallholder farmers from developing countries and other economically deprived societies (Wani et al. 2009). In the Guinea-Savannah zone of West Africa, rainy-season farming is a major source of income for the rural population (Cooper et al. 2008; Müller et al. 2013). Climate change projections suggest an increase of sub-seasonal variability of the frequency and intensity of rain events across the Guinea-Savannah (Morris et al. 2009, Sylla et al. 2016a, Salack et al. 2016); this variability could in turn affect agricultural production through reduction of water availability (Iglesias et al. 2007).

In this article, we analyze the agriculturally relevant climate metrics across the Awun river basin in Nigeria. This basin is located in Guinea-Savannah part of West Africa (Fig. 1) where many farmers are facing challenges related to rainfall variability as they are mainly involved in rain-fed agriculture. Within this region, high inter-annual variability of onset and cessation of the rainy season leads to a lack of skillful predictability using traditional knowledge (Efe 2009) and challenges in determining the optimal crop selection and planting date each year (Mugalavai et al. 2008). Generally, yields may suffer significantly with either a late onset or early cessation of the growing season and a high frequency of damaging dry spells within the growing season (Mugalavai et al. 2008; Garcia-Carreras et al. 2015). The ability to accurately estimate the actual start of the season therefore is vital for farmers (Mugalavai et al. 2008).

Fig. 1
figure 1

Map of Kwara State showing Awun basin in gray. The position of Ilorin synoptic station and the box used for simulations extraction are represented

Full size image

Several studies have tried to assess local rainfall indicators for agriculture in Nigeria focusing on present-day climate (Odekunle et al. 2007; Marteau et al. 2009; Sawa and Adebayo 2011; Ufoegbune et al. 2012). However, in the context of climate change, emphasis need to be made towards future climate as well. This can bring insights on future climate conditions and help for short- and long-term planning in agriculture.

Previous studies on rainy season characteristics which include future climate simulations over West Africa are often done using regional technics and large-scale observational products (Mariotti et al. 2014; Klutse et al. 2016; Diallo et al. 2016) instead of local methods and data. However, regional methods are not always consistent with local methods which are calibrated locally and represent better the experience on the local scale (Fitzpatrick et al. 2015). For example, Fitzpatrick et al. (2015) identified an existing disharmony between the regional and local onset date estimation over West Africa. Moreover, these studies on projected rainfall characteristics have, often, used satellite or re-analysis data as observations to validate models’ simulations instead of local station data. This is another concern for characterizing local scale rainfall as large scale products have limitations. For instance, when comparing satellite-gauged merged observations products, Sylla et al. (2013) identified significant differences among them. Important discrepancies were also found between gauge-based observations and satellite products by Nikulin et al. (2012) and Diallo et al. (2013).Therefore, rigorous validation of climate model runs using high quality observations remains a challenge for regional (West Africa) estimation of rainfall characteristics (Diallo et al. 2016).

Understanding farmers’ perceptions on changes of agro-climatic indices has been less studied but plays a key role for anticipating the impacts of climate change (Simelton et al. 2013; Codjoe et al. 2014). It can help explore the nature of meteorological evidence of the perceived changes for better adaptive planning in future climate.

The current study attempts to fill these gaps by evaluating not only present climate but also future trend of three rainfall characteristics called onset dates, the seasonality index (SI) or spread of rainy season, and the hydrologic ration (HR) or degree of dryness of an area. To achieve that, future climate simulations are chosen carefully and corrected with local station observations instead of satellite products. The unique approach in the current study relays particularly on the use of (i) local gauge-based observations to correct model simulations, (ii) farmers’ perceptions to validate present climate rainfall characteristics, and (iii) comparatives evapotranspiration models to strengthen the generalization of results.

The study area is Awun river basin, Kwara State, Nigeria. The entire basin is located between latitudes 8° 00′ and 9° 00′ North and longitudes 4° 30′ and 4° 90′ East (Fig. 1). The watershed has an area of 954 km2. Awun basin falls, mainly, within the agro-ecological zone C of Kwara State. Zone C encompasses five local government areas namely: Moro, Ilorin West, Ilorin East, Ilorin South, and Asa. The southern part of the Basin is located in Oyo state. Annual rainfall in the basin varies from 1000 to 1500 mm. The average temperatures are 27.5 °C (wet season) and 33.5 °C (dry season). The water of the basin runs into Niger River channel. The cultivated land area within the basin is about 296 km2 (Sule and Alabi 2013).

2 Data

2.1 Regional climate model description and experiment

The regional climate model simulations used in the current study have been validated and used in previous studies exploring projected changes in extreme events and shift in climate zones over West Africa (Sylla et al. 2015; Sylla et al. 2016a). In this article, we used the latest version regional climate model of the International Centre for Theoretical Physics (RegCM4; Giorgi et al. 2012) simulated at 0.25° × 0.25° horizontal resolution. The planetary boundary layer scheme and the convective scheme are from Holtslag et al. (1990) and Emanuel (1991) respectively. The resolvable scale precipitation was treated with the subgrid explicit moisture scheme (Pal et al. 2000). The ocean flux scheme was taken from Zeng et al. (1998), while land/atmosphere interactions were described using the Community Land Model version CLM3.5 (Oleson et al. 2008). RegCM4 simulations were run for 20 years (1985–2004) for the reference period and 20 years (2080–2099) for future scenarios at 25-km horizontal grid spacing over the domain selected by Browne and Sylla (2012) (i.e. 5S/25N and 25W/25E) and considered as optimal for regional climate modeling domain for West Africa. The two global climate models (GCMs) namely MPI-ESM-MR (Max-Planck Institute Earth System Model-Medium Resolution; ~ 1.9° × 1.9° horizontal resolution; (Giorgetta 2012a, b, c)) and GFDL-ESM-2M (Geophysical Fluid Dynamics Laboratory Earth System Model version 2M; ~ 2.5° × 2.0° horizontal resolution; (Dunne et al. 2012, 2013)) were used to provide initial and boundary conditions. The resulting simulations are respectively referred hereafter MPI_RegCM4 and GFDL_RegCM4 in the results sections. The model time step is 6 hourly with boundary conditions and sea surface temperature taken from the driving GCMs. The choice of GCMs was justified by their relatively good performances over Africa (Elguindi et al. 2014). For future scenarios, the new storylines of the Intergovernmental Panel on Climate Change Fifth Assessment Report (Moss et al. 2008) called “Representative Concentration Pathways” (RCPs) were used for each GCM-driven runs. The medium, or some mitigation, scenario (RCP4.5) and the extreme or business as usual scenario (RCP8.5) were used for the future 2080–2099 period.

2.2 Observations and RegCM4 simulations data

The RegCM4 simulations were extracted considering the field mean over a boxFootnote 1 (Fig. 1) that contains Awun basin in order to get a point-simulated data representative of the whole basin. In the meantime, Ilorin International Airport synoptic station data (8.44° N, 4.49° E; Fig. 1) was used as observations. Ilorin station is the only reliable station located inside the basin and is assumed to be representative of Awun basin. Rainfall recorded by that station is considered to be spatially homogenous over selected cultivated lands given its location in the middle of zone C (Matthew et al. 2017). Simulations include the reference period with daily rainfall (P), relative humidity (RH), and minimum and maximum temperatures (Tmin and Tmax) for the period 1985–2004 and for MPI and GFDL-driven runs. The future climate also includes the same variables for the period 2080–2099 and for the two forcing scenarios (RCP4.5 and RCP8.5). Synoptic station-based observations for similar variables were purchased at the Nigerian Meteorological Agency, Abuja for the period 1985–2014. Observations for a period of 20 years (1985–2004) were used to correct the model simulations and those for the period 1995–2014 were used to characterize present climate.

2.3 Bias correction of regional climate model (RegCM4) simulations

Most climate models exhibit systematic errors (biases) in their outputs (Themeßl et al. 2012).Footnote 2 Model biases can be caused by, among other factors, limited horizontal and vertical spatial resolution. This is because the spatial resolution is related to the parameterized convection schemes mentioned earlier. Bias correction methods are used to overcome these errors. The linear scaling described first in Newton et al. (2012) was applied to bias-correct the simulated P, Tmin and Tmax, and RH. The advantage of this technique is its ability to correct means values (Teutschbein and Seibert 2012). Gauged-based observations and model runs for the period 1985–2004 were analyzed to compute scaling factors which were subsequently used to correct future model runs. The methodology is described in the Appendix.

3 Methods

3.1 Onset of the growing season

The onset definition first presented by Benoit (1977) and widely used by many studies in Nigeria (Ati et al. 2002; Edoga 2007; Laux et al. 2008; Mugalavai et al. 2008) was adopted in determining onset of the growing season. Benoit (1977) defined the onset date for northern Nigeria as the date when accumulated daily rainfall exceeded 50% of accumulated potential evapotranspiration (ETp), i.e., ∑(P − 0.5 × ETp) > 0 for the remainder of the season, provided that no dry spell longer than 5 days occurs immediately after that date. This onset definition was chosen instead of many others (Olaniran 1984; Stern et al. 1981; Marteau et al. 2009; Sivakumar 1988) because it includes potential evapotranspiration rather than based on rainfall only. Benoit (1977) onset definition seeks to ensure that the rain compensates water losses by evapotranspiration throughout the growing season. After the onset, the soil will keep enough moisture and protect crops from any plausible long dry spells which may occur during the growing season. ETp was calculated using two models namely Blaney Morin Nigeria (BMN) (Duru 1984) and Hargreaves and Samani (HS) (Hargreaves 1994) to ensure that onset dates computed do not depend on the limitations of a particular ETp model. These models were chosen instead of the recommended Penman-Monteith equation (Allen et al. 1998) for their reliability in calculating ETp in Nigeria when there are limited climatic variables. Each model requires only three climatic variables to calculate ETp (Ilesanmi et al. 2014). According to BMN model, daily potential evapotranspiration can be calculated as in Eq. 1:

$$ {\mathrm{ET}}_p={r}_f\left(0.45{T}_{\mathrm{mean}}+8\right)\left(520-{\mathrm{RH}}^{1.31}\right)/100 $$
(1)

Hargreaves (1994) formula is defined in Eq. 2 as

$$ {\mathrm{ET}}_p=0.0135\;\mathrm{KRS}.\mathrm{Ra}.\sqrt{T_{\mathrm{max}}-{T}_{\mathrm{min}}\cdot \left({T}_{\mathrm{mean}}+17.8\right)} $$
(2)

In these equations: ETp (mm/day) is the daily potential evapotranspiration; rf is the ratio of monthly radiation to annual radiation; RH (%) is the daily relative humidity. Tmean (°C), Tmin (°C) and Tmax (°C) are the daily mean, minimum, and maximum temperatures respectively. Ra (mm/day) is the extra-terrestrial radiation. KRS is the radiation adjustment coefficient. Hargreaves (1994) recommended using KRS = 0.16 or “interior” regions and KRS = 0.19 for coastal regions. Interior regions (case of Awun basin) indicate inland areas far from the coasts.

3.2 Seasonality index (SI)

SI was computed for the present and future climate. It measures the spread and steadiness of the rainfall during rainy season. Walsh and Lawler (1981) have categorized SI in several classes described in Table 1. The table shows the ranges of SI with corresponding rainfall regimes and was used to identify rainfall regimes or characteristics within Awun basin. Walsh and Lawler (1981) mathematically expressed seasonality index SI following Eq. 3.

$$ \mathrm{SI}=\frac{\sum_{n=1}^{12}\mid {\overline{X}}_n-R/12\mid }{R} $$
(3)

In this equation: R (mm/year) is the mean annual rainfall and \( {\overline{X}}_{\boldsymbol{n}} \) (mm/month) is the mean rainfall of month n.

Table 1 Seasonality index (SI) classes and rainfall regimes (Walsh and Lawler 1981)
Full size table

3.3 Hydrologic ratio (HR)

HR is the degree of wetness or dryness of a place. It is defined as the ratio of mean annual rainfall to ETp (Adefolalu 1988; Sawa and Adebayo 2011). The value indicates soil moisture deficiency or surplus. It is an accurate method of estimating water availability considered as soil moisture and thus appropriate drought indicator. The index helps in agricultural decision making as it provides information on the best areas for a particular crop to thrive and give high yield. Considering land ownership constraints, HR can actually aid farmers to choose, in the area where they are allowed to grow, the crop types that will sustainably and successfully grow. The lower the value derived, the drier the wet season. According to Adefolalu (1988), when HR = 1, then optimum crop yield exists. Table 2 shows the HR values with corresponding ecological zones in Nigeria.

Table 2 Hydrological ratio and ecological zones (Adefolalu 1988)
Full size table

3.4 Farmers’ perceptions

A stratified random sample of 100 local farmers who grow locally common crops (maize, sorghum, yam, and cassava) within the Awun river basin was chosen to study their perceptions on present climate rainfall characteristics. Farmers, across communities, were asked about their perceptions on the trends of onset/cessation, the length of the growing season (LGS), and the frequency of dry spells over the past 20 years. The survey concerned local farmers who had, at least, 20 years’ experience in farming. This corresponds to the period 1995–2014 that is also used to calculate present-day agro-climatic indices over the basin.

The current study did not integrate how farmers and scientists agreed on the definitions of the different agro-climatic indices. It investigated whether farmers’ perception on indices trends has a scientific or meteorological evidence only. Further data collection is required to investigate the understanding and connection between local/indigenous knowledge and scientific knowledge on these indices.

3.5 Statistical analyses

Statistical significance tests were performed for the trend and change in mean onset dates over the present and future scenarios. The Mannkendall test analysis of linear trend significance was carried out. This method has been used in many studies across the globe (Partal and Kahya 2006; Obot et al. 2010; Manzanas et al. 2014). Welch t test, often called unequal variance t test, was used to test the change in mean onset dates.

4 Results and discussions

4.1 Present-day onset, cessation, and growing season length

As discussed in Section 3.1, two comparative HS and BMN ETp schemes are used to help calculate monsoon onset dates. Figure 2 presents the observed trend of onset dates for the present climate 1995–2014. Onset dates vary from 1 year to another. The best fit lines are negative (y = −1.197 × + 140.77) and (y = − 2.2303 × + 148.97) for HS and BMN respectively implying Julian days. This is also confirmed by the negative tau values (i.e., − 0.233 and − 0.295 for HS and BMN respectively) of the significance test (Table 3). This indicates a decreasing Julian days and implies a progressive early start of rainfall for the present period over Awun basin. However, p values of 0.1625 and 0.0743 were found for HS and BMN respectively implying that only the present-day onset date trend under BMN model is significant at 90% confidence level. Mean onset dates based on Benoit (1977) definition are presented in Fig. 4 for the present climate (1995–2014). The start of the growing season begins in 6th May and 5th May using HS and BMN models respectively for the present climate. The estimations are close to those obtained by Olaniran and Sumner (1989) using the same onset definition. However, onset dates show some differences with previous values that were obtained in Kwara State using only rainfall-based models (Olanrewaju 2012). This is because local onset dates are known to differ depending on the definition chosen and dataset used (Fitzpatrick et al. 2015).

Fig. 2
figure 2

Trend of present climate (1995–2014) onset dates using both HS and BMN ETp models. The square dot lines indicate the 3-year centered moving average to smooth the year to year variability

Full size image
Table 3 Trend significance test using Mann-Kendall test for the present and future scenarios. Positive and negative tau values indicate the increasing and decreasing trends respectively
Full size table

Farmers’ perceptions on rainfall characteristic findings are given in Fig. 3, and each characteristic is categorized as (i) normal, (ii) early, and (iii) late for onset and cessation. The length of the growing season (LGS) and the frequency of dry spells are categorized as (i) normal, (ii) increased, and (iii) decreased. According to Fig. 3a, b, the majority of farmers have noticed that onset (59% responses) and cessation (49% responses) start early over the last 20 years. Following Fig. 3c, the largest responses (49% responses) agreed on the decrease of the LGS over this period. Figure 3d also indicates that the largest share of farmers (49% responses) think that the frequency of dry spells is likely to increase in the area. Therefore, farmers’ perceptions on the onset date trends are consistent with scientific findings as presented in Fig. 2 for the last 20 years. The modeled onset trends in the present climate (1995–2014) is therefore supported by local farmers’ knowledge as being early.

Fig. 3
figure 3

Farmers’ perceptions on rainfall patterns over the last 20 years

Full size image

4.2 Projected onset dates

The projected future change in mean onset date is presented in Fig. 4 alongside with current present-day values. Under RCP4.5 scenario, mean onset dates are 2nd June and 13th June for MPI and GFDL-driven runs respectively when using HS model. With the same scenario, the onset dates are 23rd May and 3nd June for MPI and GFDL-driven runs respectively when using BMN model. This corresponds to 4 weeks delay roughly in onset dates for the scenario with some mitigation (i.e., RCP4.5). For RCP8.5, the projected mean onset dates are 30th June and 20th June for MPI and GFDL-driven runs respectively using HS model, while with BMN model, the mean onset dates expected to be 14th June and 5th June for MPI and GFDL-driven runs respectively. This corresponds to about 8 weeks delay in onset dates under the business as usual scenario. Also, changes in mean onset dates between present-day and all future scenarios are significant at p < 0.05 (Fig.4). Thus, onset dates will occur within a range of 4 to 8 weeks later than currently experienced. The expected delay of onset dates in comparison to present climate is therefore clear. These results, based on high-resolution regional climate simulations, confirm the findings of Woodley (2011) who also projected a late onset in this part of Nigeria under the optimistic (B1) and extreme (A2) scenarios at the end of the twenty-first ccentury.

Fig. 4
figure 4

Mean onset dates calculated for the present climate (1995–2014) and future scenarios (2080–2099) using HS and BMN ETp models. Asterisks indicate significance test of changes in mean onset dates from present to future scenarios

Full size image

The projected trend of onset dates is shown in Fig. 5a, b for the future climate (2080–2099) using two ETp models. Onset dates fluctuate from 1 year to another as in present-day climate. The best fit lines show a small positive (the equations are shown on Fig. 5a and b) for all scenarios, independently of the ETp models used, with the exception of GFDL_RegCM4_RCP4.5 which shows a negative slope. The same positive and negative patterns are also confirmed with tau-values of the significance test (Table 3). Therefore, we interpret that Julian days are likely to increase in the future with the exception of GFDL_RegCM4_RCP4.5. The increase is further confirmed with the ensemble means (i.e., Ensemble_mean_RCP4.5 and Ensemble_mean_RCP8.5) presented in Fig. 5. However, only the trend of Ensemble_mean_RCP8.5 scenario is significantly increasing at 95 and 90% under the HS and BMN models respectively (Table 3). This increase implies a likely progressive late start of the growing season for both future scenarios in Awun basin. We conclude that onset progressively starts earlier in recent times but will start late towards the end of the twenty-first century in Awun basin, especially, under the business as usual scenario.

Fig. 5
figure 5

Trend of future climate scenarios (2080–2099) onset dates using HS (a) and BMN (b) ETp models. MPI_RegCM4_RCP4.5, GFDL_RegCM4_RCP4.5, MPI_RegCM4_RCP8.5, and GFDL_RegCM4_RCP4.5 refer, respectively, to the scenario with some mitigation (RCP4.5) driven by MPI, the scenario with some mitigation (RCP4.5) driven by GFDL, the business as usual scenario (RCP8.5) driven by MPI, and the business as usual scenario (RCP8.5) driven by GFDL. The ensemble mean for each scenario is also presented. The 3-year centered moving averaging is presented for each scenario

Full size image

There are some possible explanations of the future changes in onset dates. One of the reasons is related to changes in temperature. There is a strong evidence regarding the projected increase of temperature over West African unlike precipitation which is more uncertain and does not show a drastic change in the future (Rowell 2012; Orlowsky and Seneviratne 2012; Barros et al. 2014; Sylla et al. 2016b). Thus, future changes in onset dates found are expected to be largely driven by temperature changes since Benoit onset date definition has the advantage of taking into account evapotranspiration in future scenarios.

4.3 Seasonality index and hydrologic ratio for present and future climate

The results of SI analysis are summarized in Fig. 6. Under the present-day climate, the SI is 0.78, which is classified as seasonal (around 6 months of wet season) according to Walsh and Lawler (1981) (Table 1). This classification is logical since the Awun basin is located within the Wooded Savanah climatic zone (Adefolalu 1988). In the future and for the RCP4.5, the SI values are 0.85 and 0.89 for MPI and GFDL-driven runs respectively. Under the business as usual scenarios, the values are 0.87 for both MPI and GFDL-driven runs respectively. These values of SI range between 0.80 and 0.99. According to Walsh and Lawler (1981), that range corresponds to a “markedly seasonal with long drier season.” The interannual variability of SI presented on Fig. 7 using boxplots confirms the increasing SI values from the present to future scenarios. For instance, the interquartile range and central value (e.g., median and mean) of SI for future scenarios are greater than present-day climate. Therefore, all future scenarios of rainfall regimes will shift from seasonal type of 6 months wet season to a shorter wet season type.

Fig. 6
figure 6

Mean seasonality indices (SI) for present-day and future climate

Full size image
Fig. 7
figure 7

Interquartile range of seasonality indices (SI) for present-day and future climate. MPI4.5, GFDL4.5, MPI8.5, and GFDL4.5 refer, respectively, to the scenario with some mitigation (RCP4.5) driven by MPI, the scenario with some mitigation (RCP4.5) driven by GFDL, the business as usual scenario (RCP8.5) driven by MPI, and the business as usual scenario (RCP8.5) driven by GFDL

Full size image

As for the HR or degree of dryness, the average HR in the region is 0.74 and 0.73 for the present climate using HS and BMN ETp models respectively (Fig. 8). This result confirms that Awun basin is located in the wooded savannah zone but close to the rainforest area as categorized by Adefolalu (1988) (Table 2). The future scenarios project an important shift of this degree of dryness. For instance, under the RCP4.5, the degree of dryness is 0.62 and 0.59 for the MPI and GFDL respectively using BMN model while, under the RCP8.5, the hydrologic ratios are 0.55 for both MPI and GFDL-driven runs. Also, the interannual variability of HR, presented on Fig. 9 using boxplot for the two ETp models, confirms the decreasing from the present to future scenarios. Central tendency values (i.e., median and mean) of HR for future scenarios are less than those of the present-day climate. It is thus likely that HR will decrease in the future, and the decrease will be accentuated under the business as usual scenario. Since the lower the ratio the more the region is drier (Adefolalu 1988), our findings imply that Awun basin will get drier than it is today. The findings confirm that climate trend of Kwara State is towards more aridity (Olanrewaju 2012; Sylla et al. 2016a, b).

Fig. 8
figure 8

Mean hydrologic ratio (HR) for present-day and future climate using HS and BMN ETp models

Full size image
Fig. 9
figure 9

Interquartile range of hydrologic ratio (HR) for present-day and future climate using HS and BMN ETp models. MPI4.5, GFDL4.5, MPI8.5, and GFDL4.5 refer, respectively, to the scenario with some mitigation (RCP4.5) driven by MPI, the scenario with some mitigation (RCP4.5) driven by GFDL, the business as usual scenario (RCP8.5) driven by MPI, and the business as usual scenario (RCP8.5) driven by GFDL. _H and _B indicate HS and BMN respectively

Full size image

5 Conclusions

In this analysis, we assess changes in mean and trends of onset dates, SI and HR at the end of the twenty-first century over the Awun basin using outputs from high-resolution regional climate model simulations considering a scenario with some mitigation (i.e., RCP4.5) and a business as usual world (i.e., RCP8.5). It is found that onset of the growing season will be late compared to present-day for both scenarios in Awun basin, but the delay will be severe in the RCP8.5 scenario. The delay is projected to range from 4 to 8 weeks in the future. In addition, the rainy season will get shorter for the RCP4.5 but much shorter for the business as usual scenarios. Finally, Awun basin will get drier than it is today under both scenarios. These results highlight the importance of social, agricultural, and genetic developments to make the farmers of Awun basic resilient to climate risks. Although the study used projections from a single climate model driven by two global climate models instead of ensemble projections encompassing several regional climate models to facilitate the uncertainty assessment, the results can still be considered to some extent with regard to forewarning and decision-making support in agriculture.

At local level and for present climate, crops sowing should be encouraged in May when the start of the growing season is set. Kwara State agricultural extensions should provide farmers with more precise early warning predictions, for instance, seasonal forecasts whose skills have increased recently over West Africa (Ingram et al. 2002). In the future, there is a need for an integrated planning between governmental agricultural agencies and local farmers to prepare for the forthcoming shortening of the rainy season and increase of the dryness. This planning can include provision of water information services, based on climate forecasts and projections, to support agriculture. In this way, farmers can take better decisions about specific crop types to use, planting and harvesting dates, fertilizer applications, and water allocation optimization for irrigation. More research is also needed to integrate scientific and local knowledge systems to produce useful and usable predictions (culturally accepted) for farmers. More broadly, Nigerian government and governments from other countries together with private sector partnership should encourage practices (agricultural, industrial, and domestic practices) which will not increase greenhouse gasses emission up to the level of RCP8.5: clean and environment-friendly technologies for both supplied energy and consumption could be used.