*2.1.3 HadGEM2-ES global climate model*

The most important tool in predicting the future climate is the modeling of climate [11]. Climate modeling studies have been conducted to determine the effects of climate changes that may occur in future periods. In Turkey, climate modeling studies were conducted within TSMS, and the final results were shared in 2015. In this study, the data on the selected meteorological parameters related to the 20-km-resolution climate projections were used on the basis of the report from HadGEM2-ES global model data and the RCP8.5 scenario used in "Turkey Climate Projections with New Scenario's and Climate Change (TR2015-CC)" by TSMS [3] for Turkey and the neighboring region. The RCP8.5 scenario is the scenario with the highest predictive radiation forcing and greenhouse gas concentration. In other words, RCP8.5 expresses the most pessimistic condition for the future periods. In this scenario, the radiative forcing reaches 8.5 w/m2 in 2100 and continues to


#### **Table 1.**

*Parameter matrices used in the least squares method.*

increase after 2100. HadGEM2-ES is a second-generation global model developed by the Hadley Center, a research organization affiliated with the UK Met Office [12].

### **2.2 Method**

#### *2.2.1 Regression analysis*

In this study, province-based regression equations were established using the least squares method (LSM) with sunflower yield values from the 29 provinces from 1985 through 2014 and the 7 selected climate parameters. After that, the potential impact of climate changes that are projected for the future periods (2016–2040, 2041–2070, and 2071–2099) on yield of sunflower has been put forward with using the generated high-rate regression equations and climate projection data (**Table 1**).

In the study, the regression analysis equation created by LSM was as follows:

$$\mathbf{y} = \mathbf{A}\mathbf{s} + \mathbf{B}\mathbf{p} + \mathbf{C}\mathbf{h} + \mathbf{D}\mathbf{k} + \mathbf{E}\mathbf{t} + \mathbf{F}\mathbf{m} + \mathbf{G}\mathbf{v} + \mathbf{H} \tag{1}$$

where the dependent variable *y = yield. A, B, C, D, E, F, G, and H* are coefficients, and the independent variables were as follows:


*v* = number of days with daily minimum temperature ≤−5°C

### **3. Results and discussion**

This study was conducted in order to determine the possible effects of climate change on sunflower yield in 29 provinces where intensive sunflower cultivation has been conducted in Turkey. The periods of 2016–2040, 2041–2070, and 2071–2099 were determined as future periods.

**29**

**Figure 2.**

*Possible Impacts of Climate Change on Sunflower Yield in Turkey*

of predicting province-based yield were very high.

**3.2 Province-based sunflower yield change projections**

*Results of province-based yield meteorological parameter regression analysis.*

yield were periodically examined.

*3.2.1 Years 2016–2040*

*3.2.2 Years 2041–2070*

the 29 provinces (**Figure 3**).

the 29 provinces (**Figure 4**).

**3.1 Province-based yield-meteorological parameter regression analysis**

In the first part of the study, it is aimed to determine the quality of the relationship between the variables using regression analysis. According to the results of multiple regression analyses using LSM between climate factors and yield, the rates

The results indicate that predictability was between 0.38 and 0.50 in three provinces (Karaman, Kırşehir, and Uşak), it was between 0.51 and 0.69 in 11 provinces (Afyon, Aydın, Çorum, Diyarbakır, İstanbul, Kayseri, Kocaeli, Kütahya, Sakarya, Tekirdağ, and Tokat), and very high predictability (0.70–0.91) was determined in 15 provinces (Adana, Aksaray, Amasya, Ankara, Balıkesir, Bilecik, Bursa, Çanakkale, Edirne, Eskişehir, İzmir, Kırklareli, Konya, Osmaniye, and Samsun)

In the second part of the study, multiple regression equations were used with the 20-km-resolution climate projection data from the HadGEM2-ES global climate model and RCP8.5 scenario. Yield estimation analyses were conducted for 2016–2040, 2041–2070, and 2071–2099. The obtained results were compared to the average yield values of 1985–2014, and the changes that may occur in sunflower

The results of the analyses conducted for 2016–2040 using the climate projections data in the regression equations predicted that sunflower yield would increase in 16 of the 29 provinces, decrease in 12 of the 29 provinces, and no change in 1 of

Obtained results of the analysis conducted for 2041–2070 using the climate projections data in the regression equations, it is predicted that there will be an increase in sunflower yield in 17 of the 29 provinces and a decrease in yield in 12 of

*DOI: http://dx.doi.org/10.5772/intechopen.91062*

(**Figure 2**).
