Anticipating heavy rainfall: Climatological aspects

Thirty years of climatological data were used to analyze heavy rainfall events. These data revealed several distinct patterns associated with heavy rainfall events. Comparing the daily values to the 30-year means, it was found that significant heavy rainfall events are often associated with large departures from the 30-year climatological data. Fields such as the specific humidity, temperature, wind, and heights were examined. Significant anomalies in each field were present in many record, or near record rainfall events. Specific humidity departures of 2-4 standard deviations above the 30-year mean appeared to be the best predictant of large excessive rainfall events.

Several record rainfall events will be shown from the climatological data set including the record flood event of 19 January 1996. During this event, 850 hPa specific humidity values were in excess of +5 standard deviations above the 30-year mean. Similar events, which did not produce nearly as much rainfall and flooding, showed a similar overall pattern however, specific humidity anomalies were on the order of +1 to +3 standard deviations above the 30-year mean.

The great northeast ice storm and warm episode of 7-10 January 1998 also revealed +3 to +4 specific humidity anomalies over New England. Unlike the 19 January 1996 event, the anomalies were oriented in an east-west direction. This overrunning event produced heavy rains and significant ice accumulation in the low-level cold air across northern New England and southeastern Quebec.

Cases were classified by type to reveal anomaly patterns associated with heavy rainfall events. These data show great promise in forecasting the patterns associated with significant rainfall events. Using model grid files in real-time compared to the 30-year climatological data may offer a method to identify potential heavy rainfall events.