Topography Level

[Marquez Feliciano]

Regionalmodel sensitivity simulations in which the height of elevated terrain was reduced to explore simulated changes in features of the low-level jet (LLJ) are presented. Such an approach has not been reported, and it provides complementary insight to the previous LLJ studies. The simulations were carried out for a 45-day period during the 1993 summer flood in the central United States, when strong LLJs were frequent. The simulations illustrate directly the significance of topographical blocking, leeside cyclogenesis, and terrain thermal effects exerted by the Rocky Mountains in support of LLJ formation. In particular, it is shown that in the absence of topography the ridging from the Bermuda high extended considerably westward with weaker southerly flow over the High Plains, thus diminishing the potential for LLJ development. The slope-induced nocturnal horizontal thermal gradient was indicated to have a significant role in the formation of the LLJ.

As implied by the aforementioned studies, the present consensus is that the most important forcing mechanisms for the LLJ are the response to diurnal boundary layer evolution over slopes and elevated terrain and the dynamical relationship to upper-level flow. It has to be emphasized that these two types of processes must not be viewed as competing or mutually exclusive explanations for the LLJ. Rather, the characteristics of the LLJ are established by their combined influences. General discussions of the combined forcing mechanisms have been given by Mitchell et al. (1995) and Zhong et al. (1996), among others.

Although both theory and observation have extensively documented the Rocky Mountain slope effects on the LLJ [see Strensrud (1996) for a review], direct quantification of the contribution of the elevated terrain to the LLJ forcing, through elimination or modification of the topography in regional model simulations, has not been reported to our knowledge. A hypothetical modification of the topography in the continental United States in a model simulation will affect the characteristics of the LLJ that are topographically forced. Adopting such an approach would provide complementary insight into the knowledge obtained in previous LLJ studies. In particular, it would outline a reference meteorological state absent of the topographical forced processes. Alteration of the topography in a model simulation is likely to modify the following LLJ forcing mechanisms: (i) the background southerly flow in the south-central United States that prevails during summer as a result of modification of the summer meteorological systems over the region, (ii) leeside cyclogenesis in the eastern slopes of the Rocky Mountains, and (iii) the along-slope induced nocturnal thermal gradient and the related mesoscale ageostrophic flow component that contributes to LLJ formation.

The flood of 1993 occurred at the maturity of an El Nio event with moderate SST anomalies over the equatorial Pacific. The above-normal convective heating in the intertropical convergence zone (ITCZ) shifted toward the equator (Trenberth and Guillemot 1996). This positive phase of the North Pacific teleconnection pattern persisted over the Pacific for 4 months prior to the flood. Then the ridge that prevailed over the western United States diminished and thus the zonal flow in the western Pacific was intensified. This zonal flow provided a channel for cyclones propagating directly to the central United States.

Synoptic patterns over the United States during the flood were characterized by a weaker Bermuda high, deeper leeside troughs east of the Rocky Mountains, and more frequent occurrence and stronger intensity of the LLJ (Mo et al. 1995; Arritt et al. 1997). The stronger upper-level jet streams and associated storm tracks pushed southward while the subtropical high retreated along the ITCZ.

Global model simulations. Global predictability vanishes after a couple of weeks, so such an approach is not well suited to the study of the details of seasonal regional processes relevant to a given anomalous large-scale environment such as the one that prevailed during the summer of 1993.

Regional climate model forced continuously in time by observed meteorological lateral boundary conditions, so as to maintain the large-scale atmospheric conditions of 1993 in the modified topography simulations. In sensitivity simulations we found that if lateral boundaries are not too close or too far from the center of the domain this approach would provide a reasonable compromise for the modeling evaluation. Specifically, we carried out simulations while extending the domain farther to the east, south, west, and north to evaluate the final placement of the lateral boundaries.

In locations where topography was reduced, the initial conditions and lateral boundary conditions in the resulting void were extrapolated as follows: standard atmosphere lapse rate for the temperature and constant (same as at the surface) for wind and relative humidity. In the present simulations the lateral boundaries were located far away from the area of interest (the LLJ region east of the Rocky Mountains), minimizing spurious effects of altered terrain. Most parts of the lateral boundaries are at sea level or associated with very shallow topography; only small portions of the lateral boundaries are over high terrain (in Central America, and northwest Canada). Thus, overall the prescribing of lateral boundary conditions when topography is removed is likely to have only secondary spurious effects.

In this subsection we present simulated lower-atmosphere patterns implying potential LLJ characteristics. It would be difficult to present the average flow during LLJ events for the simulated period, because of the spatial and temporal variation of the LLJ from one case to another. Therefore we present the simulated 45-day averaged geopotential and flow, which imply the potential for LLJ development.

Relatively strong daytime southerly flow is a prerequisite for the development of a pronounced nocturnal LLJ. At 1800 UTC (around local noon), the difference in the flow field east of the Rocky Mountains between CTRL (Fig. 2c) and FLAT (Fig. 2d) is noticeable. This result illustrates that the effects of elevated topography on the intensification of the daytime flow potentially are strongly conducive to nocturnal LLJ formation.

The 45-day average difference in wind velocity (CTRL minus FLAT) at σ = 0.91 at 0600 UTC showed a pronounced cyclonic flow perturbation over the western United States (Fig. 3). At its eastern side, this perturbation is forced by the combined effects of topographical flow blocking and leeside cyclogenesis. The flow perturbation is further supported by thermally induced pressure perturbation over the elevated terrain in the control simulation [see observational evaluation in Reiter and Tang (1984) and Tucker (1999)]. The southerly flow component difference was enhanced east of the Rocky Mountains in the area where the LLJ was simulated (see Fig. 5 later). The presented relative vorticity at 850 hPa (smoothed) corresponding to the velocity difference between the CTRL and FLAT simulations provides another perspective on the characteristics of the cyclonic perturbation.

In an attempt to gain additional perspective of the LLJ forcing we established, based on observations, two dominant synoptic classes over the Rocky Mountains and their eastern side (following T.-C. Chen 2003, personal communication): (i) days in which cyclonic/trough systems (denoted C) prevailed in the lower atmosphere over the region (20 days) and (ii) days in which anticyclonic/ridge systems (denoted A) prevailed over the region (15 days). The remaining 10 days were not sufficiently well defined to be included in either class. Average simulated geopotential height at 850 hPa and flow at σ = 0.91 at 0600 UTC are presented in Fig. 4. For class C in CTRL (Fig. 4a), a trough was centered over a large segment of the eastern upper slopes of the Rocky Mountains with a strong south-southwesterly flow east of the trough line (the flow was stronger than the CTRL averaged for all simulated days; see Fig. 2a). When topography was removed (Fig. 4b), the troughing was weaker (as forcing for the leeside cyclogenesis is absent). The geopotential gradients and flow at the typical location of observed maximum LLJ in Oklahoma and northern Texas also were weakened. The corresponding simulated patterns for class A in the CTRL simulation indicated ridging from the Bermuda high was well defined to the east of Rocky Mountains and troughing over the northwest United States (not shown). The flow at the LLJ location was southerly, but weakened compared to the C simulations. In FLAT, the absence of blocking effects of the Rocky Mountains caused the Bermuda high to expand considerably westward and northward and the flow weakened significantly over the observed LLJ maximum location.

In this note we presented sensitivity simulations that explored simulated changes in LLJ features in the central United States when terrain elevation was reduced. Such an approach has not been reported, and it provides complementary insight to prior studies of LLJ. The integration period was a 45-day window selected from the 1993 flood when strong LLJs were frequent. Evaluations were made in terms of time-averaged properties during the simulation period.

3a8082e126