Time Series Regression Analysis on Effects of Climatic Change and Variability on Production and Yield of Bean, Sweet Potato and Cassava Crops in Kenya
There is a much uncertainty in the prediction of production and yield for cereal and tuber crops in Kenya. These uncertainties can only be addressed and mitigated with the aid of accurately developed parsimonious statistical models which encompasses several important predictive variables. In this paper the effect of climatic factors and rural population on the production and yield of important food crops grown by poor households in Kenya are investigated. Secondary time series data on climate change factors, population, crop production, and yield were used in the study. The data was recorded from 1961 to 2020, obtained from (FAO, 2020) and (NASA, 2020). In addition, temperature and other climatic data were collected from Climate Change Knowledge Portal (2020). Time series regression model building approach was used in the analysis. First, all the variables were tested to find if they were stationary. In case the data was non-stationary, transformation methods of differencing, logarithm or the ratio method was applied. The study found that there was a strong correlation(r=0.9017) between the rural population and the carbon dioxide emission. In the case of sweet potato, the four significant factors that affected production were floods (p-value=0.006), rural population (p-value=0.000), area under harvest (p-value=0.002) and drought (p-value=0.031). The sweet-potato-regression model had the highest predictive power, 82.4 percent. For beans, only one factor, the area under harvest (p-value=0.000) was significant; while for cassava, the only two significant factors were: area under harvest (p-value=0.001) and the cube-area under harvest (0.000). The significance of the area under harvest being a major factor pointed to lack of technology adoption by the small household farmers. The study recommends that farmers should plant larger acreages of crops, employ technology, and adopt measures that can mitigate against incidences of drought.