Abstract
Archived radar reports, derived from the National Weather Service radar network, were used to estimate the average annual frequencies of thunderstorm days across New York State for the period 1978–81. The archival records consist of manually-digitized radar (MDR) data, available on magnetic tapes and arranged as hourly, digitally-encoded radar reflectivity values within a high-resolution grid of reporting blocks, each 45 × 45 km. Analyses of these data made use of an experimentally-derived relationship between radar reflectivities and the presence and intensities of thunderstorms. The radar-based thunderstorm day climatology generally agreed to within 15% of conventional, surface-based thunderstorm day statistics reported for the same period by National Weather Service (NWS) offices located within range of two or more network radars in the State. Poorest agreement between annual totals was found at selected NWS offices in the Greater New York City Metropolitan Area and northward into the lower Hudson River Valley, in far western New York and over far northern New York. Where redundant, near-continuous network radar coverage was available, a northwest-to-southeast increase of thunderstorm days, approaching an annual maximum of 45 in downstate New York was revealed. This gradient in thunderstorm day activity is significantly different from that depicted on isokeraunic maps derived from conventional thunder observing protocol. Because the MDR data are archived on a high-resolution grid of reporting blocks, local thunderstorm maxima on a scale of tens of kilometers may be resolved. Analyses further revealed that 5–25% of all thunderstorm days contained sufficiently vigorous storms to be characterized as “intense”. The greatest frequency of intense thunderstorm days, approaching 8 annually, was located in the highly-populated region of the State along the New York-New Jersey borders, northwest of the Greater New York City Metropolitan Area.
