Construction contracts let you claim time for "abnormal" or "unusually severe" weather. None of them — AIA A201, the FAR, ConsensusDocs — define what that means. So contractors argue about rain with adjectives and usually lose. The data is out there, but buried and cumbersome for some to wrangle while under stress. I built Ilystics to put a number on it: ten years of NOAA daily station observations run through a soil drying model, producing how many workdays a site actually loses. Not rain days — workdays, which is the unit contracts are written in. The thing I didn't expect: in 26 of the 45 metros, the wettest month isn't the worst month. Boston averages 4.28" of rain in July and loses 8 workdays; January averages 3.69" — less rain — and loses 16. Double the loss on less rainfall. The reason is the drying tail. Winter rain lands spread out on cold ground and each event starts its own multi-day clock before equipment can work. Summer storms cluster and their drying periods overlap. So a bid priced off rainfall totals systematically misprices winter earthwork. What's free: baselines for all 45 metros, the per-year distribution behind every average, and a PDF exhibit for any US ZIP with no account. The whole dataset is CC BY 4.0 on Kaggle and HuggingFace — I'd rather people check the numbers than trust me. Happy to answer anything about the model. The methodology is public because a number an opposing expert can't verify is worth nothing in a dispute.

Ilystics

Construction documentation for weather delays simplified

Construction documentation for weather delays simplified