EXTRAPOLATION TO THE 50-MB. LEVEL FROM 100-MB. DATA IN ANTARCTICA

Abstract

Winter-month data obtained at United States Antarctic stations during the ICY are used to compute coefficients for regression equations relating the 100-mb. temperatures and heights to the 50-mb. heights. It is found upon trial that the layer thickness can be estimated as accurately from the 100-mb. temperature alone as from both the temperature and 100-mb. height; therefore, the simpler equations in only one variable are used. The average errors in computed heights are less than 2 dekameters and the extreme errors 6 dkm. or less when tests are performed on independent data. A measure of the average lapse rates within the layer between 100 mb. and 50 mb., derived from the same data as the equations mentioned above, proves to be as good a predictor as any of the regression equations. These mean lapse rates show a regular variation with both latitude and month. The largest values (greatest instability) occur in the coldest months and the highest latitudes, and the greatest layer stability in lower latitudes and warmer months. Additional 50-mb. heights are extrapolated from the observed 100-mb. data for all of the stations considered for radiosonde ascents reaching 100 mb. but not 50 mb.; means of the computed heights are compared with the means of the observed 50-mb. heights to determine if any bias in the observations can be demonstrated.