Abstract

A series of two-dimensional (2-D) numerical experiments has been conducted to examine the effects of differential surface heating on flow over a dry, 2000 km-wide plateau. Two effects, found by Benjamin and Carlson in three-dimensional simulations to be significant in the regional severe storm environment, also occur in these 2-D experiments. These effects are a diurnal variation in the intensity of the lee trough and the development of a low-level inversion downstream as the mixed layer, which developed over the hot plateau, is advected over potentially cooler air.

When the plateau is strongly heated and surrounded by lowlands with no surface heating, the leeside pressure trough intensifies by an extra 1–3 mb. Subsequently, the low-level flow ahead of the lee trough also increases by several meters per second when surface heating is allowed. The diurnal modulation of this feature suggests that the low-level moist flow toward regions of potential convection in cases such as those modeled by Benjamin and Carlson will tend to be strongest in the late afternoon and early evening. It is shown that this effect is primarily due to the superposition of a plateau heat low upon the mountain wave circulation. To a lesser extent, differential vertical mixing of momentum between the deep mixed layer and surrounding regions also tends to enhance the lee trough. This differential mixing momentum mechanism is active in the presence of an isolated, deep mixed layer and moderately strong lower tropospheric flow even if there is no elevated terrain.

The development of the elevated mixed layer inversion appears to depend more strongly on a horizontal gradient of soil moisture and surface heating than it does upon a gradient of terrain elevation. However, while such an inversion may be produced by differential advection and differential heating in the absence of terrain, it will be stronger and develop more rapidly in a shearing environment if the strongly heated region is also elevated.

Abstract

An objective analysis scheme for meteorological variables on constant potential temperature surfaces is presented. The analysis uses a form of multivariate statistical interpolation and is designed th retain mesoscale detail in disparate observations including rawinsonde, surface, aircraft, satellite, and wind profiler data while combining them with a forecast background (first guess) field. The wind and mass field analyses are interdependent. The horizontal correlation of forecast error on isentropic surfaces is modeled with an analytical function from statistics collected for this study; the vertical correlation of forecast error is modeled as a function of potential temperature separation. These correlations determine the weights applied to observed-minus-forecast increments in the analysis. The analysis is two-dimensional except with respect to single-level data where it is three-dimensional.

Comparisons of isentropic and isobaric analysts are shown, and examples of the effects of single-level (aircraft and surface) observations on isentropic analyses are presented. Although variable in space and time, these datasets are often of higher density than the rawinsonde network, and they support increased resolution of mesoscale features in the analysis. More importantly, the examples reveal that three-dimensional analysis increment structures, especially in the vicinity of fronts, appear to be more physically reasonable in an isentropic analysis than in an isobaric analysis.

Abstract

A short-range numerical prediction model, which is part of a real-time 3-h data assimilation and forecast system, is described. The distinguishing feature of the model is the use of terrain-following (σ) coordinate surfaces in the lower troposphere combined with isentropic (θ) surfaces aloft. Such a hybrid coordinate system allows modeling of processes in a convectively unstable boundary layer while retaining tile advantages of θ coordinates in representing upper-tropospheric frontal and jet-stream structures. The hybrid approach used in this model represents a in major departure from previous hybrid formulations in atmospheric models, oven though it has been used for more than ten years in oceanic modeling. Part I of this two-part paper contains a thorough description of the model and the results of validation experiments. Results of North American case studies wig be reported in Part II.

Abstract

A comparison was made of temperature and wind observations reported by rawinsonde and Aircraft Communications, Addressing, and Reporting System (ACARS)-equipped commercial aircraft separated by less than 150 km in distance and 90 min in time near Denver, Colorado, in February and March 1992. Only data made on aircraft ascents and descents reported automatically were used. A total of 4440 matched data pairs were obtained for this period. The sample was analyzed in total but also as a function of time and distance separation, height, time of day, and ascent versus descent. A standard deviation temperature difference of 0.97°C and rms vector wind difference of 5.76 m s⁻¹ were found for the entire sample but were reduced, respectively, to 0.59°C and 4.00 m s⁻¹ when only data pairs separated by less than 25 km and 15 min were used. The study provides an upper limit to the combination of rawinsonde and ACARS observation and reporting errors and mesoscale variability, but it is not possible to distinguish the exact contributions from each of these sources. However, overall these statistics indicate that the rawinsonde data used were more accurate than that reported in a previous study and that the accuracy of the ACARS data was somewhat higher still.

This paper examines the influence of data from the NOAA Wind Profiler Demonstration Network on a mesoscale data assimilation system. The Mesoscale Analysis and Prediction System is a 3-h intermittent data assimilation system configured in an isentropic-sigma framework. To measure the impact from profiler data on 3-h forecasts valid at 0000 and 1200 UTC, parallel runs with and without profiler data were verified against rawinsonde data. A sample case study is also presented to show the magnitude of the modifications at verification sites. In evaluations from case studies and statistics gathered over longer test periods, the profiler data improved the overall short-range forecasts in the study area. This improvement was most evident at 300 hPa where the root-mean-squared wind errors (averaged over the verification area) were reduced by 0.7 m s⁻¹, and corresponding height errors were reduced by 2 m. The 300-hPa improvement in short-range forecasts from the case study at individual rawinsonde stations was as large as 10 m s⁻¹ for winds and 40 m for heights.

Abstract

An assimilation system is presented that was designed to provide timely, detailed, and coherent analyses of surface data, even when the data are collected in rough terrain where station elevations differ widely and observations are often subject to local effects. Analyses with improved spatial continuity are obtained from these data through careful choice of analysis method and variables. The analysis method has the ability to handle varying data density, and the analysis variables, when possible, were chosen in such a way as to cancel out the effects of elevation differences. In addition, the method accounts for physical blocking and channeling by mountainous terrain by incorporating elevation and potential temperature differences in its horizontal correlation functions. The correlation functions also enable the method to move accurately represent surface gradients.

An hourly analysis cycle is used in which each analysis uses as a background the previous hourly analysis (a 1-h persistence forecast). The cycling is important in providing temporal continuity between analyses.

Detailed explanations of the analysis variables and method are given, along with a discussion of the objective quality-control procedures necessary to ensure reliable analyses in an operational environment. The assimilation system has been used experimentally by National Weather Service forecasters since 1996. Quality-control statistics summarizing the observational errors of surface stations across the 48 contiguous states are also presented.

The effects of variable terrain on the analyses are demonstrated in examples. Sample analyses are presented, including diagnosed fields, for a severe-storm case. Overall, the surface analyses described here allow better temporal and spatial resolution than the current operational National Meteorolegical Center surfaces analyses.

Abstract

A combined longwave and shortwave radiative transfer model was used to determine effects of Saharan dust on the radiative fluxes and heating/cooling rates in the atmosphere. Cases are treated for cloud-free and overcast conditions over the ocean and for cloud-free sky over desert.

A benchmark comparison, made for the cloud-free ocean case between our calculations and those from Wiscombe’s detailed model, yielded results which were in close agreement. For moderately heavy dust amounts commonly measured over the Sahara and the eastern tropical Atlantic Ocean, typical calculated aerosol heating rates for the combined longwave and shortwave spectrum were in excess of 1 K day⁻¹ for all three cases for most of the atmosphere beneath the top of the dust layer (500 mb). For the ocean case, maximum heating rates are found near the level of maximum concentration (700 mb), and also near the surface beneath the Saharan air layer (below 900 mb).

Net fluxes determined at the top of the atmosphere for the ocean cloud-free case were very insensitive to changes in dust optical depth. For the cloudy oceanic and desert cases, the reflectivity of the earth-atmosphere system diminished with increasing dust optical depth and approached that for the ocean case at large optical depth. In all three cases, the dust reduced the downward radiative flux into the ocean or desert while at the same time it increased the heating in the atmosphere, thus indicating a stabilizing effect by dust on the temperature lapse. However, further speculation concerning climatological significance of these results must be tempered by a need for further study of interactions between aerosol heating and atmospheric circulations, and between aerosols themselves and cloud microphysical processes.

Abstract

A set of observation system experiments (OSEs) over three seasons using the hourly updated Rapid Refresh (RAP) numerical weather prediction (NWP) assimilation–forecast system identifies the importance of the various components of the North American observing system for 3–12-h RAP forecasts. Aircraft observations emerge as the strongest-impact observation type for wind, relative humidity (RH), and temperature forecasts, permitting a 15%–30% reduction in 6-h forecast error in the troposphere and lower stratosphere. Major positive impacts are also seen from rawinsondes, GOES satellite cloud observations, and surface observations, with lesser but still significant impacts from GPS precipitable water (PW) observations, satellite atmospheric motion vectors (AMVs), and radar reflectivity observations. A separate experiment revealed that the aircraft-related RH forecast improvement was augmented by 50% due specifically to the addition of aircraft moisture observations. Additionally, observations from en route aircraft and those from ascending or descending aircraft contribute approximately equally to the overall forecast skill, with the strongest impacts in the respective layers of the observations. Initial results from these OSEs supported implementation of an improved assimilation configuration of boundary layer pseudoinnovations from surface observations, as well as allowing the assimilation of satellite AMVs over land. The breadth of these experiments over the three seasons suggests that observation impact results are applicable to general forecasting skill, not just classes of phenomena during limited time periods.

Abstract

A method for station or grid point reduction of surface pressure to sea level or some other level is presented that shows improvement over the standard reduction method in the western United States. This method (MAPS SLP-Mesoscale Analysis and Prediction System sea level pressure) uses the 700 hPa temperature to estimate an “effective” surface temperature from which the temperature of the hypothetical layer beneath the ground is estimated. The use of this “effective” temperature instead of the observed surface temperature is responsible for the improved reduction since it varies more smoothly over space and time and is more representative of the temperature variation found above the boundary layer.

The MAPS SLP reduction was compared with the standard reduction and altimeter setting reduction in statistical comparisons of geostrophic wind estimates with observed winds and in a case study. A 21-month comparison between geostrophic and observed winds was made over different geographical regions, times of day, rotation angles and seasons. The results showed that the MAPS SLP reduction performed better than the standard reduction in the western United States, but not in other regions with generally low elevation. In general, the correlation between sea level geostrophic winds and observed winds was found to be dependent on the Froude number. A statistical comparison using a smaller sample between MAPS SLP and the Sangster geostrophic wind, which is not a station reduction, showed similar skill over the western United States. The case study also showed that the pattern over the western United States was more coherent and less anomalous with MAPS SLP that with the other reductions.

Abstract

Model experiments have isolated several effects of surface heating and topography which may act in concert to focus the potential for severe thunderstorms in certain areas downstream of dry elevated terrain. These effects are examined primarily in results from 12 h and 36 h three-dimensional model simulations of the 9–11 April 1979 (SESAME I) case, but are also found in simulations of the 9–10 May 1979 (SESAME IV) case. The experiments are performed using the Pennsylvania State University mesoscale model, which includes a sophisticated boundary-layer component with horizontally varying surface characteristics and cloud effects on the surface radiation budget.

Comparisons of simulations with and without surface fluxes of heat and moisture confirm that differential surface heating, topography, and differential advection may combine to produce a stabilizing effect. In certain synoptic situations, air strongly heated over the arid Mexican plateau rides over moist potentially cooler air downstream, thereby forming a strong restraining inversion or “lid” over Texas. This elevated mixed layer inversion prevents the occurrence of thunderstorms over a large area; instead, storms are focused in the region where a southeasterly low-level flow underruns the northern and western boundary of the inversion. Model trajectories also support the hypothesis that this lateral boundary of the lid, a baroclinic zone approximately between 800 and 500 mb, is formed by confluence between the Mexican airstream and a subsided polar airstream which has traversed the upper-level trough approaching from the western United States.

Model experiments also reveal that the strength of the low-level flow toward the severe storm region may itself be enhanced by effects related to surface heating and topography. First, the lee trough which forms with flow aloft across the southern Rockies and Mexican plateau intensifies as a result of daytime surface heating. This effect, which is related to differential heating and differential mixing of momentum, is significant in that the low-level flow east of the lee trough from the Gulf of Mexico toward the severe storm region is also strengthened by several meters per second. Secondly, transverse circulations associated with jet streaks, which were crucial in inducing low-level flow in both the SESAME I and IV cases, were also found to intensify in the presence of surface heating.