WMOD00079

• Contract/Permit #: #98-1
• Applicant: Western Kansas Groundwater
• Project Name: Kansas Weather Modification
• Date: 1/1/1998
• Weather Modification - Doc Type: Report

\
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
\
I I
I

V. HAIL DISTRIBUTION

As usual, we will look at the insurance crop-hail damage claim numbers in a variety of ways
in an attempt to assess what has happened in the current year regarding hail damage in the WKWMP.
However, we first need to discuss why we use these numbers:

Crop-hail damage claims are sent to us during the season by the Kansas/Oklahoma Hail Loss
Service based in Washington, Kansas. Data from these sheets provide us with the location of the of
the damage (by township, range and section) and the date and type of crop damaged, but nothing
about the value of crop losses. Statisticians would prefer having actual dollar numbers relating to the
amount of crop insured (liability) versus the amount paid out by insurers (losses) and how well the
insurance premiums covered their losses. Unfortunately, such meaningful cropchail insurance statistical
data related those things are not available to the public until around the following April.

We keep track of the hail claims data for the entire state during the season, categorizing them
by date of reported occurrence, by county and for areas inside and outside our WKWMP target area.
It is a crude means of attempting to assess hail damage and it usefulness has limits; however, in the
past we have obtained advanced insight suggesting trends in hail damage long before 'official'
insurance data are made known about the most. recent season. Most inferences we make after
analyzing the hail claims damage numbers does have some usefulne.sS. For instance, when we wanted
to find out what should be our optimum period of operation for a season we used the historic claims
numbers and found what appeared to be the first and last significant day of hail damage in an average
season, so now we base our yearly operational period around those dates. We think these data also
can give us some advanced insight into how effective our expanded fleet of aircraft is doing to mitigate
hail falling from the higWy destructive supercell storms which wreak tremendous crop and property
damage on us as they pas; through our area. From these data we can. to compare one hail-day to
another to determine relative success in seeding, although such comparisons may not be foolproof as
we occasionally learn. One such case occurred when Stevens County made a brief one-year return to
the WKWMP in 1996. Stevens County had the most hail claims of any county in SW Kansas in 1996,
727 claims. We thought Stevens County was going to show disastrous losses when the dollar results
were made known. However, when the actual dollar losses and liabilities are examined, we found out
Stevens County had its lowest loss-to-liability (loss-cost) value in 1996 when comparing it to the high
damage, 5-year period 1993 through 1997. Similarly, Stevens County's 1996 loss-cost was well below
its average loss-cost dating back 21 years to 1977. All of this suggests a high likelihood there may
have been a beneficial seeding effect despit~ the high numbers of claims.

Other unknowns during a year are how many farmers buy insurance and, of the total crop
grown, how much of it is insured? If relatively few farmers insure their crops in a county, that county
might show relatively little crop-hail damage instead of relatively high damage when high damage
might have been the case, etc. Another drawback to using crop-hail damage numbers is that the date
given for the hail occurrence occasionally is incorrect. For instance, this past summer there were 15
days out of the reported 64 hail-days (during the regular season) for the entire target area in which
there were no operations, or 23% of the days. For those 15 days 74 claims were recorded, or 3% of

23