description abstract | Recent developments in meteorological objective analysis have been directed towards improving numerical model initialization, often at the expense of preserving detail in data rich regions. The objective scheme presented here is designed to give detailed, yet computationally efficient analyses over data rich regions and to be independent of any numerical model influences. The results, which accurately preserve observed data and gradient information, are particularly suited to diagnostic studies. The analysis scheme uses both observed station data and interstation gradients to interpolate pressure, wind, and moisture information from vertical cross sections onto isentropic surfaces using cubic polynomials. Three-dimensional grid point mass and moisture analyses at specified isentropic intervals are obtained by first constructing a series of nearly parallel, north?south vertical cross sections over an area of interest and then filling in the space between adjacent vertical analyses using a second set of nearly orthogonal cross sections. The choice of isentropic coordinates allows the objective analysis to use detailed information included within the vertical structure of sets of individual, quality-controlled soundings to capture significant horizontal thermal gradients. Gradients obtained from the mass analyses are then combined with observed wind data to obtain observed wind analyses which have spatial variations consistent with the scales of variations noted in the detailed mass fields. Examples are given which illustrate the detail and continuity inherent in the technique when applied to radiosonde data obtained at a variety of scales. | |