To contribute to the predictions of short-term heavy rainfall, we extracted the cases of heavy rainfall in Tokyo from high-density observation data and analyzed the characteristics of the surface wind convergence that preceded the rainfall. The error in divergence due errors in the surface wind observations was evaluated, and the correspondence between the divergence and vertical flow near the ground was examined.
An index―“effective convergence”―the increase in the amount of convergence that precedes heavy rainfall was defined using the same method that was used to determine the effective rainfall, and the temporal change in the effective convergence was examined. An increase in convergence was observed in about 25%-40% of cases of heavy rainfall. In addition, the greater the effective convergence before the heavy rainfall, the greater is the cumulative and peak precipitation. There was a good correspondence between the areas where there was an increase in the effective convergence and the heavy rainfall. As a result of using the effective convergence to make predictions of heavy rainfall, it was found that the effective convergence reached the threshold value at least 60 mins earlier than the occurrence of rainfall in half of the cases where heavy rainfall was detected. Although the false alarm rate was high, it was suggested that monitoring the effective convergence would be useful in the predictions of short-term heavy rainfall.
This report studies intense rainfalls over Nagasaki area (northwestern area of Kyushu District) on 23 July 1982 using precipitation data at about 90 stations of Japan Meteorological Agency (JMA data) and other organizations (non-JMA data). Maximum 1-hour precipitation and its occurrence time over Nagasaki area are examined to confirm the time-spatial consistency of the observations. It is also shown that the standard-deviation of the maximum 1-hour precipitation (120-150 mm/h) in data rich area of about 7 km radius is about 10 mm. Analysis of 1-hour precipitation including the non-JMA data revealed detailed variations of intense rainfall. The consistency of the 10-min precipitation data could not be examined because of the insufficient spatial density of the observations. The10-min precipitation data show a few peaks of 10-min precipitation within the strong rainfall period.