CCM Question 2 Key

CCM Exam Question 2

A homeowner in the town of Greenbrae (3 Corte Cayuga), in Marin County, California has hired you to:

(a) Determine the range of seasonal rainfall amounts she can expect at the property most (67%) of the time;

(b) Determine the 100-yr 1 hour event at her site;

Part (a)

Since areal variability related to meteorology is small in California, except that due to topography, the consultant used the variability and return period information for the NWS rainfall site in Kentfield, approximately 2 km distant from the site, as representative for those which would be observed at the site. The consultant took the standard deviation for Kentfield's rainfall and calculated coefficient of variation (16.30/46.96 = 0.346). He then multiplied that by the estimated mean annual rainfall at Greenbrae from Rantz's map (~36") to get +/- 12.49". That range was added and subtracted to get the rainfall range that could be expected 67% of the time, 23.52 to 48.49".

Parts (c) and (d)

The ratio of Greenbrae's estimated average seasonal rainfall to that at Kentfield as 0.77. The consultant then multiplied this ratio to the 100 y 1 h and 24 h rainfall amounts that were obtained from California State Department of Water Resources Depth-Duration-Frequency tables for Kentfield to to obtain: 100 y 24 h ~ 9.20" and 100 y 1 h ~ 2.29".

Use of return period information should be made with great caution. The concept of return period is often misused. In order for return period information to be somewhat useful and valid for meteorological information at least a 30-year sample of information should be obtained. That is because only for a 30-year or longer period of record for rainfall information will the statistics calculated for that sample be somewhat representative of the same statistics calculated for the longer record, if it existed. As an example, there is about a 64% probablity that a 30-yr record will actually contain a true 30 yr storm (as a storm is defined, for any duration). There is only a 26% probability that a 30-yr record will actually encompass the 100 yr storm (from Bruce and Clark, 1966, p. 130).

Even if the return periods could be interpreted in a straightforward manner, the term 100 yr storm does not mean that such an event would only occur once in a century, but that over a very long period, say 500 years, such a storm would occur 5 times, on average. In essence, there is a 1% chance that such an event would occur in any given year. Even so, such a concept is grounded in the assumption that climate has not changed. Thus, the term "100 y" event, as appear above, appear in the statistics, but really have little DIRECT meaning, because of the shortness of the record (Kentfield's is less than 100 years in length). However, such terms do establish a reasonable estimate of the potential "severe" event.

Figure 1: Location Map. 3 Corte Cayuga Site Location and Kentfield NWS Cooperative Rain Gauge Locations Indicated by Red Squares