COMBINATION OF INITIAL AND PROGNOSTIC DATA IN STATISTICAL CLOUD PREDICTION.

• Main Authors: Cooley,Duane S., Ostby,Frederick P., Chisholm,Donald A., Pavlowitz,Abraham M.
• Other Authors: TRAVELERS RESEARCH CENTER INC HARTFORD CONN
• Format: Text
• Language: English
• Published: 1964
• Subjects: *CLOUDS; WEATHER FORECASTING; CLOUD COVER; STATISTICAL ANALYSIS; ATMOSPHERIC TEMPERATURE; POLAR REGIONS; Arctic
• Online Access: http://www.dtic.mil/docs/citations/AD0434072; http://oai.dtic.mil/oai/oai?&verb=getRecord&metadataPrefix=html&identifier=AD0434072

A study was made of the statistical prediction of low- and total-cloud amounts from parameters derived from numerical weather prediction (NWP), Air Weather Service moisture and temperature analyses and predictions, and other meteorological variables. Cloud data and other atmospheric data over Eurasia and the Arctic were analyzed on a grid mesh of approximately 208 mi (the NWP grid). Predictability of cloud amount was evaluated by using screening regression and multiplediscriminant analysis and testing the predicitions on independent data. Predictors considered included cloud amount, 1000-, 850-, and 500-mb height, geostrophic-wind components, surface and upper-level temperatures, stability, vorticity, divergence, upslope motion, thickness of isobaric layers, condensation-pressure spread, and vertical velocity. Some parameters were used in previous tests of statistical diagnosis of clouds. Values taken from predicted fields of the parameters are used to specify clouds at the predictand time. Parameters are considered at initial time for prediction at lag. (Author)