982 CLIMATOLOGY 
counties or townships. Records are based on Weather 
Bureau observations and claims against the various 
insurance companies. There is a considerable variation 
m occurrence of damaging hailstorms from locality to 
locality. These variations are quite startling. Roth 
reports that the probability of hail damage varies so 
much that in extreme eastern Nebraska the insurance 
rate is 3 per cent of the amount insured, while in western 
Nebraska 18 per cent is charged. 
Rain insurance is generally written for the protection 
of outdoor events against losses. These include carni- 
vals, fairs, and sport contests. The insurance is paid 
when rainfall at the particular locality exceeds a mini- 
mum amount on the specified days or within the speci- 
fied hours. This amount is often fixed at 0.1 in. It 
is easy to calculate the risk for such an occurrence from 
the extensive available rainfall statistics. At present, 
the insurance companies use weekly rainfall statistics. 
The frequency of occurrence of amounts above the 
minimum during the period of record is established 
and the risk is computed. Many decades of records are 
needed for this purpose. 
Windstorms are similarly treated. In that case the 
damaging limit has to be established. Normally, wind 
speeds of less than 30 mph do not cause appreciable 
damage. However, it is not quite as easy to establish 
the frequency of windstorms above this or a similar 
limit as it is to establish the frequency with which 
rainfall exceeds a certain amount. At many stations 
winds are merely estimated according to the Beaufort 
scale. Not until recent years has there been a widespread 
network of stations equipped with recording anemome- 
ters. Even so, older instruments (such as the U. S. 
Weather Bureau triple register) are not very useful 
for determining peak gusts, because the instruments 
integrate the wind velocity over a period of time, but 
the peak gusts are the ones that cause the damage. 
In this last case, the climatologist has to use his 
meteorological judgment to supplement the records. 
Where upper-air data are available, atmospheric-sta- 
bility criteria can serve as useful guides to estimate 
gustiness. The frequency of incidence of hurricanes 
and tornadoes is also helpful. Yet the statistics are quite 
incomplete and we may find that one station which was 
affected by a severe storm has recorded winds in excess 
of gale force, while another close by has not, during the 
period of record, experienced such an unusual event. 
Adjustments of the records by analogies have then to 
be made in order to arrive at an appropriate risk 
factor. 
In areas with sparse records this procedure may also 
have to be applied to other elements for calculation of 
insurance risks. Long-record stations will give the clima- 
tologist the probable shape of the frequency distribu- 
tion. This shape, but not the absolute values, can 
then be used for the adjustment of shorter records 
at nearby stations. 
In all these cases it is tacitly assumed that the past 
performance of climate is likely to continue essentially in 
the same fashion in the future. By and large this condi- 
tion is fulfilled, but risks, once established, should be 
recalculated whenever a sufficient number of new ob- 
servations become available. 
Single Point—Simple Time Series—Multiple Climatic 
Element 
Examples par excellence for this category of problems 
can be found in the field of design and specifications 
for protection of man against his environment. Clothing 
and housing fall into this group. When we build a house 
we want it to compensate for the exigencies of climate 
so that conditions indoors approach the zone of comfort 
as much of the time as possible. This aim governs all 
design factors. Nearly all climatic elements have to be 
considered for the purpose. In a preliminary fashion, 
analyses of the effects of various weather factors upon 
architecture and construction have been made [88, 77]. 
TaBLe LV. Curmatic Factors in Housine Desien* 
> ao 
8 otal yee yy | 353 
se | Bie 3 | 32 
Climatic factors & bo 2 ie ‘ Se ‘sg 
6S |S aie. | seis 
sf| 2 | 3s | & | sal 2 
ae fs | Sie le 
Temperature xX/|X|xX|]X/} xX] X 
frequencies 
Thermal heat | Frequency of x x 
hot and cold 
days 
Degree days x xX 
Sunshine hours x x x 
Clear and cloudy xX | xX XC || XX 
Radiation days 
Solar intensity 2 | BE |} UC | OK 
Solar height OXE ||| XE || EXE 
Wind direction xX | X|]xX/ xX x 
Winds Wind speed — xX | xX x x 
Strong winds x 
Precipitation DE || 2S | 2 |] X\ 
Snowfall x xX | xX 
Excess precipi- xX BE PS || AC] Ae 
Atmospheric tation 
moisture fees! Bea 
Rainy days x | X x 
Fogs x 
Thunderstorms x 
Humidity BNC ONG XGT | EXG OX 
* Crosses (X) are entered for those elements which are of 
some importance for design of the particular feature. 
Without going into the details of the reasoning in- 
volved, we can give a schematic picture of the relation 
of various weather factors to diverse phases of housing 
construction (Table IV). 
The design of the individual features of the house will 
have to be geared to the climatic conditions. Houses are 
