A warm current arrives off the coast of Australia for just three weeks. That is enough for more snapper larvae to survive, and a year later fishermen enjoy a bountiful catch. But how do you spot such a short episode in a long series of daily water temperature measurements? In the past, scientists either scanned the data by hand or averaged over the season, losing crucial details.

The IMPIT app on Shiny R lets you upload your own data and tweak the sieve parameters. In examples with snapper, scallops, and wheat in Australia, the model with the index explained catches and yields better than raw temperature or pressure values. Now any agronomist or ecologist can quickly check which weather episodes decide the fate of a harvest or fishery.