ALBERT

Description: No abstract available.

Description: No abstract available.

Description: This paper reviews the considerations in evaluating the skill and significance of screening multiple linear regression (SMLR) models. Formulations and procedures are given along with relevant references to prior studies. Topics discussed include predictor selection, serial correlation, artificial skill, true skill, and Monte Carlo significance testing. New results with wide applicability in the assessment of SMLR model skill and significance are presented in graphical form. However, the results are restricted to situations involving predictors which are independent of one another and are serially uncorrelated. The methodology presented is suggested for use in both model evaluation and experimental design.

Description: No abstract available.

Description: No abstract available.

Description: Comparison between in situ aircraft observations of temperature and National Meteorological Center and Global Weather Central analysis fields of temperature is presented for a continental and oceanic flight route. The standard deviations of the temperature differences over several hundred flights are found to be 2.5 and 3.5°C for the continental and oceanic route, respectively. A bias towards warm temperatures of about 0.85°C for the analysis fields was found for the oceanic route. Only small differences are found between the NMC and GWC analysis field temperatures.

Description: Simple geometric considerations suggest that in the same way as broken cloud cover appears practically total near a ground observer's horizon, the anisotropy of the longwave radiation emitted to space from an area of broken cloud should be substantially enhanced. Using a simple model consisting of parallel bands of “black” clouds of rectangular cross section, imbedded in an otherwise horizontally homogeneous vertically stratified atmosphere, we evaluate the longwave radiance emergent in different directions, averaged over the cloud pattern, and find this expectation to be confirmed in a wide range of cases. At large zenith angles, the emergent radiance can exhibit considerable dependence on azimuth, relative to the direction of the band structure. Although such azimuthal dependence practically disappears for many 3-dimensional broken cloud fields (e.g., hexagonal cells), the enhanced zenith angle dependence remains. This can be a source of bias in cloud extractions or Earth radiation budget determinations based on geostationary satellite data, since zenith angle is directly related to geographical location of the point under observation. With our model cloud arrays, and considering radiative transfer in the atmosphere, we have evaluated the bias in the broad-band longwave (3–100 μm) radiant exitance values derived, assuming horizontal homogeneity, from either broad-band or narrow-band (10.5–12.5 μm) radiance data. We find that the bias for near-nadir observations can be as large as 20% (40 W m−2) in extreme cases involving bands of opaque high-level cloud over warm ground. It may remain substantial in situations which are encountered in nature. We examine the dependence of the broken cloud effects on the geometrical parameters of the field. We discuss how such effects might be detected and how they can influence determinations of Earth radiation budget components and of cloud cover. We note that “brokenness” will be a factor in cloud-radiation feedback.

Description: The spatial and spectral characteristics of the GOES visible and infrared images are examined. From that analysis, a scheme is developed which identifies and separates the following classes: clear skies/no snow cover, clear skies/snow cover and clouds. Clouds are then classified as high or low broken clouds and overcast. The scheme is tested for various weather situations. Comparison of the classification results with reports from ground synoptic stations and maps reflect an average accuracy of approximately 72%, and a higher accuracy (∼87%) when high or low broken clouds and overcast are considered as one class (i.e., clouds). The differentiation between clouds and snow, or no snow-covered ground has been found to be very satisfactory, even in cases of temperature inversions.

Description: No abstract available.