I.Introduction
The Ross Ice Shelf Air Stream (RAS), which flows adjacent to the Transantarctic Mountains of Antarctica (Figure 1), is a common feature of daily weather over the far eastern regions of Antarctica. Numerous cold katabatic tributaries plunging down the valleys of the Tranantarctic Mountains from the 3 km high Antarctic Plateau feed the massive river of cold dense air. The dramatic merger of the katabatic winds into the RAS is clearly evident on the satellite image displayed in Figure 1. As the RAS flows along the Ross Ice Shelf it eventually moves out over the open water of the Antarctic Ocean where it absorbs large amounts of heat from the water’s surface playing a key role in the general ocean/atmospheric circulation of the Earth. Along its path, the strong flows and sometimes turbulent and cloudy interactions with ambient air masses wreak havoc with field operations at the McMurdo research station. Forecasting the associated weather is an important challenge facing the operational forecasting community in support of field operations (Turner et al., 2000). Understanding the interaction and its role in the Earth’s climate is equally important to the climate research community.
The existence of barrier jets such as the RAS have been a subject of study for some time (O'Connor et al. 1994, Parish 1982, Colle and Mass, 1995). They have been sometimes explained by blocking effects occurring when low Froude number flow approaches an obstacle. Recently, explanations have also involved the concept of topographically trapped Kelvin-like waves that act to damn cold air against a mountain barrier (Colle and Mass, 1995). In each off these cases, however, the flows being explained consisted of cold airflows interacting with a mountain range, such as an arctic system plunging southward along the east slopes of the Rockies. RAS is somewhat unique in that it appears that the sir stream is formed of cold katabatic flows seeping through passes in the mountain range that fosters traps the RAS.
The RAS is a very shallow, narrow air stream whose microscale complexities have not been fully resolved by mesoscale modeling. Although some aspects of the RAS seem to exist in the regional scale simulations, these models generally do not simultaneously capture the katabatic flows together with the general RAS. High resolution attempts to numerically simulate the katabatic flows have been made in two-dimensions (Parish and Cassano, 2003), and some attempts have been made to simulate certain microscale interactions with barriers associated with the RAS.
Figure 1: Advanced Very High Resolution Radiometer thermal infrared image showing the Ross Ice Shelf Air Stream (RAS). Also noticeable are tongues of katabatic flow descending down from the Transantarctic Mountains. The cold katabatic winds appear relatively "warm" because of the influence of mixing on the surface temperature. The bulk temperature of the katabatic flow remains colder than the surroundings. Image is courtesy of David Bromwich at the Byrd Polar Research Center, Ohio State University.
To capture the behavior of the katabatic flows, and to represent the blocking effects of the steep slopes of the Transantarctic Mountains, a competent model of highly stratified airflow around very steep obstacles is essential. Many numerical models employ approximations that make them skillful at stratified atmospheric flow systems, so long as the topography is moderate or skillful at simulating flow around obstacles so long as the flow is not highly stratified or vertically structured. The University of Wisconsin-Nonhydrostatic Modeling System (UW-NMS) (Tripoli, 1992) is uniquely developed to do both. Its skill has been demonstrated already in simulating the 3-dimensional wind flow of the RAS around Ross Island (Seefeldt et. al., 2003).
This proposal is to study the interaction between the entire system of katabatic flows with the RAS and how they determine its behavior and structure under a variety of synoptic conditions through numerical simulation of the process. The goal will be to develop a conceptual understanding of: (a) the vertical structure of the RAS, (b) the microscale horizontal structure of the RAS, (c) the maintenance mechanism supporting those structures through heat and momentum transport, (d) how the structure is affected by ambient synoptic conditions that can impact the strength of katabatic flows, regional geostrophic pressure gradients and so on.
Below is a brief description off the UW-NMS, and its topography scheme that is key to the success of this study. Next a description of the proposed research goals is given. The results of previous NSF research relevant to this project are discussed. Finally, the broader impacts educational outreach associated with this project are discussed.
II.University of Wisconsin-Nonhydrostatic Modeling System (UW-NMS)
The UW-NMS is a nonhydrostatic, quasi-compressible, enstrophy conserving model formulated in the non-Boussinesq framework. This three dimensional, fully scalable numerical weather prediction model is capable of multiple two-way interactive grid nesting with moveable inner grids. The predictive variables of the NMS include the three components of velocity, exner function, ice-liquid water potential temperature, total water specific humidity, and five categories of precipitating hydrometeors (rain, pristine crystals, rimes mature crystals, aggregated crystals, and graupel). The UW-NMS uses a cloud active long and short wave radiation scheme (developed by: Panegrossi and Ackerman), a modified Emanuel cumulus parameterization, and Louis (1979) surface layer. The UW-NMS uses a diffusion scheme based on turbulent kinetic energy (TKE) that can be run with or without a nonlocal turbulence component.
One of the unique characteristics of the UW-NMS is that topography is handled through a variable stepped method. This differs from the more common terrain following or stepped topography used by other models. A schematic of the variable step topography is given in Figure 2a-b. With the variable step topography the lowest grid box has variable depth to exactly match surface elevation. Thin surface boxes are finite differenced implicitly to insure stability. The step boundary conditions conserve vorticity and momentum. This treatment of topography handles even the subtlest topography (elevation changes as small as one meter can be accounted for), while having no slope restrictions to severe topography.
III. Research Goals
The primary research goals of this project are:
- Successfully model the 3-dimensional structure of the RAS.
- Investigate the sensitivity of the RAS to the katabatic flow, Siple Coast confluence zone, and synoptic scale pressure gradient.
- Investigate the interaction of the RAS with localized topography around Ross Island.
Observations of the RAS are due mostly to surface observations collected by Antarctic Automated Weather Stations (AWS). Therefore the vertical and spatial structure of the RAS has not been adequately captured. Understanding the complete 3-dimensional structure of the RAS would allow for more thorough quantitative estimates of mass, momentum, and heat transfers associated with the RAS. The UW-NMS will be run using a nested grid system, where the second grid will have a horizontal resolution of approximately 1km (Figure 3). A third grid, with a horizontal resolution on the order of hundreds of meters, will be placed in areas of steep topography, as interest delegates, to further capture features like a specific tongue of katabatic flow or the flow around/over a specific barrier.
In order to determine the success of the modeling efforts, they will be compared to observations obtained by the High Latitude Integrated Sounding System (HISS). The HISS will include a Doppler sodar capable of wind measurements 15m-600m above the surface, and a Radio Acoustic Sounding System (RASS) capable of temperature measurements from approximately 300m to 2km. (For a more thorough description of the HISS capabilities please refer to proposal by Cohn.) The high vertical and time resolution observations that the HISS will provide will allow for verification of the UW-NMS. Simulations performed by the UW-NMS prior to deployment of the HISS will use the Automatic Weather Stations (AWS) and satellite imagery to verify model results. In addition to the AWS already available in the Ross Ice Shelf area, additional stations set up to support RIME activities could be used for UW-NMS verification.
It is intended that enough investigative simulations of the RAS will be conducted prior to RIME that the results may be useful for field operations during RIME. It is our goal that a successful modeling of the RAS will give insight into what flight paths would be most beneficial to the NCAR C-130, in terms of capturing observations of the key components of the RAS. Modeling of the Ross Ice Shelf wind field by the UW-NMS could also be useful to RIME related flights and observation systems by providing an approximation to the 3-D background state of the boundary layer above the Ross Ice Shelf.
The sensitivity of the RAS to katabatic flow off of the Transantarctic Mountains, synoptic scale pressure gradients, and the Siple Coast confluence zone will be examined. The persistent nature of the Ross Ice Shelf Air Stream makes studying its sensitivities complicated. The persistent nature of the RAS leads to the question: Is it possible that under different meteorological conditions the dominant forcing mechanism of the RAS may be different, but that the appearance of the RAS itself remains fairly consistent? It is anticipated that 3-D simulations of the RAS by the UW-NMS will be able to identify and quantitate the contribution of the katabatic flow, Siple Coast confluence zone, and synoptic scale pressure gradient in a variety of meteorological conditions. Previous work on the airflow across the Ross Ice Shelf, by Schwerdtfeger (1984), Bromwich (1988) and O Conner et al. (1994) suggest that the RAS is a barrier wind. While the role of the Transantarctic Mountains as a barrier is apparent, there is still room to explore the origins of the cold stable air which is unable to cross the Transantarctic Mountains, and thus creates a barrier wind. While previous work points to the influence of synoptic scale cyclones, averaged streamlines of the region (RIME science plan, 2002) show the presence of the RAS. This supports the possibility that the katabatic flow, which is more consistently present than any particular synoptic pattern could play a role in the barrier wind. This leads to the question, is it the katabatic winds that provide the cold air pool that induces the barrier wind? One way that the contribution the katabatic flow will be examined will be by examining the sensitivity of the RAS to cloud cover. The driving force of the katabatic winds is radiative cooling, which can be hindered by cloud cover. The retardation of radiative cooling, and thus possible weakening of katabatic flow will be examined in relation to the strength of the RAS. The cold air provided by the katabatic flow may induce a kelvin wave forced barrier wind, similar to that observed near the Rocky Mountains.
The varying strength of the RAS can have a profound influence on the operations at McMurdo Station. Minna Bluff, White Island and Black Island are all located upstream of McMurdo Station. In the highly stable boundary layer over Antarctica, even the most subtle of topography can influence the wind field. When approaching a barrier, if the wind field does not have the kinetic energy to overcome the potential energy associated with raising the air over the barrier, then the flow will go around the barrier. The work of Seefeldt et al (2003) showed that variation in the wind speed and stability result in the flow sometimes going over the barriers upstream from McMurdo, and sometimes going around the barriers (Figure 4).
The first year research will involve the identification of cases to study. Several cases with different synoptic patterns and cloud cover will be chosen. Initial simulations of these cases will be conducted. Also in the first year, any changes to the UW-NMS that are necessary to better study the RAS will be identified by the initial simulations and then completed. The second year of research will focus on higher resolution simulations (dx=dy=200m) of specific areas of interest, such as an individual katabatic flow out of a specific valley of the Transantarctic Mountains. By modeling a single katabatic flow on the microscale, we hope to capture an understanding of the katabatic flows interaction with the RAS at an unprecedented level of detail. We hope to capture the very fine scale turbulent structure associated with individual katabatic flow. The final year of research will work to verify the simulations, and then use the results to better quantitate transport of mass, momentum, and heat by the RAS.
IV.Results of Previous NFS Support
The PI (Tripoli) worked on "A Lake-ICE Proposal: Operations Center and Multi-scale Numerical Investigations." Through this NSF grant, Tripoli and the Tripoli group participated in the Lake-ICE field program through modeling support, scientific support, and satellite data collection support. Following the field phase, Tripoli and students performed extensive cloud resolving numerical studies of Lake-ICE observed roll convection. Results by Tripoli et al., (1999) demonstrated that multiply nested model simulations of roll convection were successful in capturing observed structures. Idealized process studies were then performed to understand the important physical processes leading to the rolls. Results from this study demonstrated a strong connection between the observed structures and local coastline geometries. This was consistent with the structures observed with volume imaging lidar (VIL). These cloud resolving simulations were significant in their attempt and success in simulating roll convection structures explicitly. The resulting simulations were some of the first to attach roll structure development to particular surface variations such as coastline geometry. Several papers have been published on work stemming from the Lake-ICE project, dealing with both observations and modeling of the boundary layer. These include: Mayor et al. (2001), Mayor et al. (2002, 2003), and Kristovich et al. (2000). The PI of this proposal (Tripoli) has submitted for publication "Numerical Study of the 10 January 1998 Lake Effect Rolls Observed during Lake-ICE" to the Journal of Atmospheric Science. A paper demonstrating that the statistical impact of roll convection can be captured with a nonlocal turbulence closure is in preparation.
The work done on the Lake-ICE grant has implications for the proposed Antarctic work. In simulating the convective boundary layer over Lake Michigan it was important to first successfully model the stable, shear driven boundary layer that approached the warm lake. The importance of the upstream stable boundary layer has led to another proposal to study further the boundary layer observed during Lake-ICE, with an emphasis on studying the upstream boundary layer (Continued Lidar and Numerical Investigations of the Internal Boundary Layer Observed During Lake-ICE). The stable, shear driven boundary layer that approaches the Great Lakes during lake effect events is akin to the stable, shear driven boundary layer observed in Antarctica. The lessons learned about the importance of subtle geometry and topography changes along the Lake Michigan shoreline foreshadow the importance that even the slightest topography will have on the structure of the Antarctic boundary layer.
V.Broader Impacts and Educational Outreach
The knowledge gained through this research will be beneficial to operational forecasting in Antarctica. Wind conditions can greatly influence flight operations in Antarctica. The cost of flight operations and support of flight operations in Antarctica is significant. The location of McMurdo Station, on Ross Island and downstream of Minna Bluff, Black Island, and White Island, makes it especially subject to topographically influences. The key understandings will facilitate improved representations of the RAS process and its attendant momentum and heat transport mechanisms in prediction models, such as those used to predict weather conditions for the McMurdo area. The results will also lead to a better understanding of how Antarctica’s local circulations are manifested on a global scale where the Antarctica katabatic circulation is known to anchor the South Polar vortex.
The goals of this proposal reach beyond pure science and a contribution to the scientific community to include public outreach. It is intended that research results will be kept on web page to encourage public interaction. The web page will not focus on just quantitative results, but will rather have a highly qualitative component that is intellectually accessible to the public. Output from model simulations will be displayed as animations made from vis5d. The focus of these animations will be to show the three dimensionality of the flow around the complex terrain. The PI and his group have a friendly work relationship with members of the Antarctic Meteorological Research Center (AMRC), a part of the Space Science and Engineering Center at the University of Wisconsin-Madison. The Tripoli research group will work with AMRC members in their K-12 educational outreach. Additionally members of the Tripoli research group have connections with GSUSA (Girl Scouts of the United States of America), and hope to bring Antarctic experience, knowledge and science to this group. It is hoped that by sharing the excitement of Antarctic research by interactions with Girl Scout groups (especially girls age 12-17), can help to make girls see science as a viable career option.