If not recorded precipitation, what data should be used? Check the methodology of predicting 500 year flood levels with only 100 years of data. With future lives at stake and millions of dollars worth of infrastructure investment at risk, any suggestions to improve the process would be welcomed.
With regard to water resources, let the planners know the exact population, ag/industrial consumption, and new uses 10, 50, and 100 years from now. They will throw that in with the new, non-historical, precipitation data and do a better job of planning.
Most of those involved in this work are licensed engineers with post graduate education and very strong statistical skills. Much of the major infrastructure in use today was designed before the current excellent computer packages were in use. Sort of like going to the moon in the sixties with slide rules as a design computation tool.
It would be great to nail the facilities response to every future event. That is the goal. Right now, we are stuck with the data in the record. Here in the states, we are witness to "climate" scientists filtering/adjusting raw data and trying to foist that on the professional community as fact. There you go. That will improve the results.