From Fig.13.7 the probability of a local white Christmas is just under 50 percent. (Because local topography, such as the effects of mountains, is ignored in Fig.13.7, the actual probability is probably not very accurately represented by the map data.)
13-QT2. Suppose the chance
for a "White Christmas" at your home is 10 percent. Last Christmas was a white
one. If for next year you forecast a "nonwhite" Christmas,
will you have shown
any skill if your forecast turns out to be correct? Explain.
Whether or not last Christmas was white, climatology predicts there is a 90 percent chance next Christmas will be nonwhite. If you predict a nonwhite Christmas and turn out correct you have done only slightly better than climatology in making that prediction, thus you have shown negligibly little skill. (If you were successful in predicting a white Christmas, then you would have manifest considerable skill.)
13-QT3. Suppose that it is
presently warm and raining. A cold front will pass your area in 3 hours. Behind the front
it is cold and snowing. Make a persistence forecast
for your area 6 hours from now. Would
you expect this forecast to be correct? Explain. Now, make a forecast for your area using
the steady-state, or trend
method.
Persistence forecast: 6 hours from now it will be warm and raining. This forecast will probably be incorrect because it ignores the effects of the frontal passage which will occur. A better (steady-state) forecast might state: "Warm and rainy, then turning colder with rain changing to snow."
13-QT4.
Since computer models have difficulty in adequately considering the effects of
small-scale geographic features on a weather map, why don't numerical weather
forecasts simply reduce the grid spacing to about one kilometer?.
There are two good reasons, time and cost. Reducing grid spacing would require greater input of data to establish initial conditions. Acquiring such data would require more observations and therefore would increase costs. If such an expanded data set of initial conditions could be acquired, calculating a forecast for the finer grid would require more calculations and therefore more computer time at greater cost.
13-PE2. When a persistent
winter pattern at 500 mb appears similar to that shown in Fig.13.24, show on the
map where you would forecast the following: (a) good
chance of precipitation; (b)
above seasonal temperatures; (c) below seasonal temperatures; (d) generally
dry weather.
See figure.
13-PE3. In Fig.13.11,
p.352, mark the position of the following cities: Cleveland, Ohio; Albuquerque, New
Mexico; and New Orleans, Louisiana. Based on the projected
movement of the surface weather
systems in Fig.13.14, p.355, make a 24-hour forecast, include temperature,
pressure, cloud cover, humidity, winds, and
precipitation (if any). compare your forecasts
with the actual weather at the end of the period in Fig.13.15, p.357.
Forecasts based upon Fig.13.14 are:
Cleveland: Slightly warmer today and tonight with light southerly winds. Clouds increasing this afternoon and evening with light snow beginning about midnight. Barometric pressure will fall throughout the period. Humidity will generally increase until near saturation is reached at the time snowfall begins.
Albuquerque: Sunny and cold today with northerly winds and very low humidity. Clear and very cold tonight. Barometric pressure will rise throughout the period.