this case and have been identified using Cook’s distances. Many of them are ignored in order to balance the regions, provinces and municipalities to be retained which is an integral element of the objectives in this study. Given the estimation procedure, this paper presents the eight models—four independent climate variable models for each data set. The results will be discussed in the succeeding section. 5. Results Having different set of samples with different set of time periods may show different trends in the effects of climate shocks as well as its covariates. It is important however, to depict the expected change in climate variables in the future to facilitate interpretation of the manifestation of changes in climate. Changes in climate Table 6 shows the general changes in the measures of climate change used in this study. It can be noted that, overall, all of the climate variables will increase in the future except for the rainfall seasonality. However, only the maximum temperature and minimum temperature will have a significant change, at 37 percent and 47 percent respectively. A closer inspection across provinces will show that there are varying changes in the inspected climate variables. Table 6. Change in climate variables 1979-2006 and 2011-2040, Philippines Climate variable Mean cumulative annual rainfall Mean rainfall seasonality Volatility of rainfall Mean 1979-2010 2011-2040 Change 2693.2 0.50 0.65 2827.7 0.50 0.67 5.0% 0.3% 3.7% Mean maximum annual temperature (degree C) 32.10 32.47 37.0% Mean minimum annual temperature (degree C) 20.16 20.63 47.0% Figure shows the provinces covered by CBMS in comparison with the country average. It shows that although Philippines is expected to have increased rainfall in the future, CBMS coverage includes some provinces that decreased in rainfall. Similarly, it can be seen in Figure9 that still, many provinces in the CBMS dataset will likely experience lower increase in minimum temperature than the country average. CBMS-FAO | 29