There are two possibilities that you get different outcomes from the author's; one is the author is wrong, second, your model specification is wrong. Where does this example come from (i.e., which book and page number)?
"15.2" and "0.04" should be replaced by "13.083" and "0.094", respectively. The author is wrong about the report of the GLM outcome. However, the possible explanation is author actually reported the outcome from some "different" models. The reference group is the low precipitation, which has 14.07 (i.e. the estimated intercept) of SO2 concentration. While the SO2 concentration is 13.083 higher for mid participation and 0.094 higher for high precipitation, respectively, comparing to the reference group.