Irrigation Requirements and Yields of Maize Crop Under Future Climate in Some Cities of Northern Cameroon

Using meteorological data obtained from Regional Model (REMO), maize yields from the years 2020 to 2099 were simulated by AquaCrop Model in Maroua, Garoua and Kaélé. These future yields are almost nil for the three cities. In view to determine the minimum quantities of water needed to improve them, the hypothesis of no water moisture stress was considered. For the four periods of 20 consecutive years (2020-2039, 2040-2059, 2060-2079 and 2080-2099), average yields of 5.21, 5.11, 4.97, and 4.73 ton/ha are obtained in Garoua, 5.05, 4.97, 4.64, and 3.87 ton/ha in Maroua and 4.91, 4.82, 4.51 and 3.69 t/ha in Kaélé. The average quantities of water irrigation (also obtained with AquaCrop) for the same periods are 13, 19, 46 and 78 mm for Garoua; 34, 48, 84 and 147 mm for Maroua and 57, 68, 111 and 171 mm for Kaélé. The yields by considering these irrigation water quantities are improved and the following values are obtained for the four periods indicated above: 5.20, 5.10, 4.99 and 4.82 ton/ha for Garoua; 5.10, 5.00, 4.78 and 4.35 ton/ha for Maroua and 4.99, 4.91, 4.75 and 4.50 ton/ha for Kaélé.


Introduction
Climate change is of concern for all the countries in the world and its consequences are related to a significant threat to the Goals achievement of sustainable development and growth in Africa (FPA, 2017;Inoussa, 2010). Future climate trends predict that large areas including the Sahel, the Horn of Africa, parts of central and southern Africa, could be warmed by about 3 to 6 °C by 2100 for an average of 4.5 °C (IPCC, 2007). It's also expected that rainfall patterns will be affected and may fall by more than 20 to 30 percent from the WMO baseline of 1961-1990(Bigot et al., 1997. Sanderson et al. (2011) projected for these regions a faster temperature rise than the global average increase during the twenty-first century. Unfortunately these predictions could be realized because according to the IPCC Fifth Assessment Report (Niang et al., 2014), an increase in the surface temperature was observed in Sahel region over the last 50 years. Also, New et al. (2006) noticed that the number of cold days and nights are decreasing. All these climatic changes will have important effects on activities such as livestock farming, hydropower generation and agriculture which are highly dependent on the climate and African countries for which, the economy is based on these activities will be greatly affected. Studies carried out in Senegal (Seck et al., 2005) and Niger (Salack et al., 2006) for example have shown that agricultural yields are experiencing drastic deficits due to the adverse effects of these climatic changes and their estimation by 2050 indicate very significant declines. Studies conducted by the IPCC (2017) indicate that by 2020, rainfed agricultural yields could fall by 50% due to soil moisture stress and 75 to 250 million people in Africa will be affected.
Several models such as WOFOST (Supit, 1994), Cropsyst (Stöckle et al., 2003), APSIM (McGown et al., 1996), DSSAT (Jones et al., 2003), EPIC (Williams et al., 1989), STICS (Brisson et al., 1998), AquaCrop ) have been designed and allow making such studies. AquaCrop model differs from others by its balance between precision, simplicity, robustness and its suitability to dealing with conditions where water is a limiting factor in agricultural production. Moreover, it simulates rainfed agriculture, additional deficit and total irrigation (Steduto et al., 2007), hydrological flow parameters. It takes into consideration the soil water content in the profile and compartments as well as the net irrigation needs , crop sequences and analysis of climate scenarios (FAO, 2020). It has been successfully used to predict yields of some culture like maize and sorghum in future water availability scenarios in Kenya (Abedinpour et al., 2014;Mwangi et al., 2019) and other countries in the world ( Olivier et al., 2016). In this work, it was used to study the maize yields in the cities of Garoua, Kaélé and Maroua, three main cities of the Northern region of Cameroon. The impacts of the climate from 2020 to 2099 on the yields of maize crops and the irrigation proposals to remedy to the situation are presented.

Study Areas
This study was carried out in the cities of Garoua, Kaélé and Maroua located in regions of Northern Cameroon. Figure 1 shows the localization of these cities on the Cameroon map and (2) Parcel All the so parameters point, perm correspond (see Table   Table 4 Vol. 12, No. 8;2020 indicators, the coefficient of determination (R²) of linear adjustment, the absolute mean square error (RMSE), the square root of the normalized mean square error (nRMSE) and the Willmott's agreement index (d) were used. Their expressions are: where, n is the total number of observations, O the observed mean values, P the predicted mean values, Oi and Pi are the observed and predicted values.
The value of the normalized mean square error (NRMSE) lower than 10% is ideal for modeling. The nRMSE values in the range of 10-20% and 20% to 30%, indicate an appropriate and moderate condition in the model predictions respectively. A value more than 30% indicates the uncertainty of the model (Jamieson et al., 1991). d goes from 0 to 1 with 0 indicating no agreement and 1 indicating a perfect agreement between the predicted and the observed data.

Simulations of the Future Yields and Irrigation Management
The simulations of future maize yields were done on one hand with all the parameters obtained from the calibration and in other hand, in the perfect condition of no water stress before flowering. For the latter, we have chosen to also determine with AquaCrop the quantities of water required for irrigation. These quantities are determined such that the exhaustion of the root zone is just above the threshold of 50% of RAW (readily available soil water).

Calibration and Validation Results of AquaCrop Model
The parameters set out in Table 5 were taken into consideration for the calibration of the AquaCrop model. The values of R 2 = 0.93 for the city of Garoua and R 2 = 0.85 for the city of Maroua were obtained by comparing the simulated yields with the observed ones (see Figure 5. These values of R 2 as well as those of RMSE, nRMSE and d all close to the ideal values for a perfect calibration of the yields (shown in Table 6), show that the relationship between the observed and simulated yield is good. The model is then capable of simulating the yields of maize crops in the studied cities.   Table 7 gi  2060 to 2 [1970][1971][1972][1973][1974][1975][1976][1977][1978][1979][1980][1981][1982][1983][1984][1985]        maize yields w n in the city of ds of maize wi n the city of K and 2042, yie his rate of irri lready sufficie the rates of ir nt differences ing irrigation a will become s ures show sim Vol. 12,No. 8; with and witho f Kaélé     In order to (2020 to 2 with those Table 8 sh and the av ions done in p he climate wi the study peri improve maize ve from one p n these cities w roua, maize e rate erage from over time as indicated above. Whatever the period studied; it is found that the values of the rate of change and the average quantities of irrigation water increase when one passes from the city of Garoua to that of Maroua and that of Kaélé. As indicated above the city of Kaélé and Maroua are more affected by these climatic changes and the effect is less significant for the city of Maroua.

Conclusions
The main objective of this work was on the one hand to study the impact of the future climate on maize yields in the cities of Garoua (9 o 18′N and 13 o 24′E), Kaélé (10 o 05′44″N and 14 o 26′37″E) and Maroua (10 o 35′N and 14 o 19′E) from Northern Cameroon and on the other hand, to estimate using AquaCrop model the quantities of water needed to improve these yields. The model was first calibrated with the data on the field from 1999 to 2004 and, by using the meteorological data of 2020 to 2099 obtained from the Regional Model (REMO) associated with the MPI-ESM global model of the RCP 8.5 scenario, the future yields were simulated. These yields were compared with those of the reference periods : 1979-2004for Garoua and Maroua and 1970-1985 Kaélé.
The results show that the yields obtained with the calibrated parameters are almost nil for the three cities. In the hypothesis of no soil moisture stress to determine the minimum quantity of water needed to remedy to the situation, we noticed that in general yields in the three cities decrease gradually from 2020 to 2099 and are higher in Maroua than in Garoua and Kaélé. These rates vary in general with rainfall, but some exceptions are observed owing the thermal increase. A comparison of simulated average yields over periods of 20 consecutive years with those of the reference periods shows that they decrease from 3 to 12% for Garoua, from 11 to 33% for Kaélé and from 8 to 30% for Maroua. The simulations with irrigation show that these yields are slightly improved in all three cities and the quantities of water needed for irrigation are less than 100 mm from 2020 to 2078 in the city of Garoua then become more important until 2099. In Maroua, the observation is similar but the quantities of water required for irrigation are greater than 100 mm from 2080 and yield remain low compared to those of Garoua. In the city of Kaélé, the simulated yields with irrigation are still lower than those of the previous cities and the quantities of water needed for irrigation are less than 100 mm until 2050 then, increase and the maximum is 266 mm. Irrigation could thus be done to improve maize yields in the three cities and combat their vulnerabilities to climate change.