Home - Presentation

Overall presentation of the project

Issued Friday 6 April 2007.


Introduction

One Antarctic climate issue that has received considerable public and scientific attention is the stratospheric ozone. While the role of polar stratospheric clouds in the depletion of ozone over the poles has been recognized for over two decades (e.g., Soloman et al. 1986; Toon et al. 1986; Crutzen and Arnold 1986), critical issues still remain concerning the chemistry and microphysics of these clouds, such as the nucleation mechanism for the nitric acid hydrate in some of these clouds (Tolbert and Toon 2001). Questions also exist regarding the interaction of these clouds with the dynamics of the stratosphere. For example, Innis and Klekociuk (2006) utilized backscatter lidar and radiosonde data to conclude that large-scale planetary waves were the drivers of the temperature fluctuations that cause these clouds with gravity waves simply modulating their fine-scale structure. In contrast, Hopfner et al. (2005) used limb soundings to show that gravity waves produced by orography could produce vortex wide polar stratospheric clouds. Shibata et al (2003) also concluded, based on a different lidar and radiosonde data set, that inertial gravity waves produced by adjustment processes associated with synoptic scale wave breaking could explain the stratospheric cloud field. The debate on this topic illustrates the wide range of scales necessary to address the issue of ozone loss and the generation of polar stratospheric clouds resulting in challenges in obtaining appropriate data sets and simulations and the debate on the role of gravity waves and the vortex.

Antarctica is an integral component of the Earth System containing 70% of our planet’s fresh water. Melting and fresh water discharges from the continent that are not balanced by increases in precipitation and storage have the potential to raise the sea levels around the globe by meters or even tens of meters. In contrast, substantial increases in snowfall and storage could mitigate or slow any sea level increases from melting of the Greenland ice sheet. Thus, a critical question for society is "Will climate change result in a significant change to the mass budget of the Antarctica ice sheet?". Unfortunately, some of the largest variances in the results of climate models are found over polar regions (e.g., Gates et al. 1996) and significant research challenges and considerable debate still exist in our current understanding current and ability to predict the mass budget and future climate of Antarctica. For example, the International Panel on Climate Change (2001) predicted that the Antarctic ice sheet would gain mass during the 21st century due to increased precipitation in a warmer climate. However, a recent international effort described by Managhan et al. (2006) concluded, based on model simulations and observations primarily from ice cores, that precipitation over the Antarctic continent has not significantly changed over the past fifty years. The inclusion of model output in determination of the observed precipitation record illustrates one of the problems in investigating this and other critical processes over the region as the sparseness of the observing network coupled with difficulty of distinguishing between snowfall and blowing snow make an accurate determination of the snowfall record difficult at best (e.g., Adams 2004). The problem of predicting future changes is complicated by the need to also address multi-scale issues. For example, intense mesoscale vortices with local convective clouds contribute to the precipitation at certain locations around the continent (e.g., Carrasco et al. 2003) and there is also a sensitivity to distant components of the global circulation, such as middle latitude blocking (Massom et al. 2003) and changes in sea surface temperatures in the tropics (e.g., Guo et al. 2004). Recent simulations with projected CO2 increases cause a poleward movement of the southern hemisphere storm track (Bengtsson and Hodges 2006).

Investigations of both the distribution of precipitation to provide insight to the mass budget over Antarctica and the issue of ozone dissipation allude to the need for improved multi-scale measurements over the region. To date, the remote, isolated location, harsh conditions and high elevations of the Antarctic plateau make it difficult to launch intensive field campaigns such as SHEBA (Randall et al. 1998) and FIRE (Curry et al. 2000), which took place in the Arctic. The Antarctica and surrounding ocean is also relatively data sparse in terms of the operational stream of in-situ atmospheric measurements, even in comparison to the Arctic. For many purposes, satellite measurements can fill such data gap over much of the earth’s atmosphere. However, even in this regard Antarctica presents unique challenges, errors and opportunities (see discussion in Comiso 2000; Bormann and Thepaut 2004). These challenges must be overcome and errors need to be reduced to produce accurate reanalyses for climate studies that are based primarily on the observed (and not modeled) conditions.

The CONCORDIASI project includes a field phase that will take place during August to November 2008 and August to November 2009. CONCORDIASI is multi-disciplinary effort based on a constellation of long duration (months) instrumented stratospheric balloons and on in-situ radiosoundings. The research foci include investigations ranging from ozone transport and depletion, the microphysics of polar stratospheric clouds, and the use of satellite data to improve operational analyses and reanalyses that provide the historical record for climate research and operational predictions over the Antarctic. These platforms will carry an unprecedented array of remote sensors and flight level sensing for meteorology, chemistry and microphysics. Through collaborations, we will advance in the study of the meteorology of the Plateau and understanding of the precipitation accumulations over Antarctica. The air from the stratosphere to the surface will be probed by the dropsondes deployed from the balloon gondolas. The measurement strategies provide a new measurement capability for polar regions. The modeling component of CONCORDIASI will be inherently multi-scale using high-resolution mesoscale models, numerical weather prediction systems and regional and global climate models. The modeling and data sets provided will result in collaborative efforts into satellite assimilation, Antarctic precipitation and the interplay between polar stratospheric clouds, gravity waves, vortex dynamics and ozone depletion.

An Overview of the CONCORDIASI Experimental Design

The international CONCORDIASI project, which is an International Polar Year (IPY) effort, will take place over Antarctica during August to November 2008 and AUgust to November 2009. CONCORDIASI is part of the IPY-THORPEX cluster of research activities that has been approved by the ICSU-WMO panel. The cluster is currently composed of ten projects with the lead researchers for these individual projects from Canada, France, Germany, Iceland, Norway, United Kingdom, and the US. We anticipate a broad sharing of ideas within the cluster so that, for example, findings made concerning the Arctic, Greenland and the Antarctic region will be rapidly and openly communicated within the cluster. Within this cluster, participation in the CONCORDIASI project includes scientists in France, United States (US), Italy, Australia, and international organizations such as the European Center for Medium Range Weather Forecasting.

Routine upper air sounding measurements are taken over Antarctica mostly along the coast, except for Amundsen-Scott and Concordia . Amundsen-Scott (US base) is located right on the South Pole and performs two radiosoundings a day. Concordia (operated by Italy and France) is located at DOME C (75S and 123E) on the plateau, and provides radiosoundings on the GTS at 12UTC. It is quite a recent site, with the first winter stay starting in February 2005 for 13 technicians and scientists. Terrestrial travel from Dumont d’Urville takes 10 to 15 days. Concordia is ideally located to validate satellite data assimilation over the continent, together with Amundsen-Scott. The plateau of Antarctica is extremely well suited for satellite assimilation since it is extremely homogeneous compared to many other environments. For sun-syncronous satellites such as Aqua and MetOp, the orbital inclination is generally around 98° and the pole itself is not directly under the satellite swath, although it is covered by the edge of swath. The main idea is to enhance the radiosounding coverage at Concordia, to bring it up to the level of twice a day, similarly to Amundsen-Scott. Furthermore, some extra radiosoundings will be taken at specific times to document local weather for more detailed precipitation studies over the plateau. In addition to the soudings, specific instruments for meteorological, snow fall and accumulations observations will be deployed at Concordia by the Institut Paul Emile Victor (IPEV), the French Polar Institut, as an important part of the CONCORDIASI project. In 2008, additional observations will be taken by enhancing the frequency of soundings at Dumont d’Urville on the coast and and at Concordia. Both will report at least twice a day from September to November 2008.

In 2009, the field activities of CONCORDIASI will be focused on a constellation of up to twenty long duration stratospheric balloons deployed from the McMurdo station by CNES (the French space agency). Such stratospheric balloon systems have been extensively utilized for atmospheric research. For instance, the GHOST campaign took place 40 years ago (Lally et al. 1966). Recently over Antarctica, a constellation of 25 CNES stratospheric balloons were deployed from McMurdo in 2005 with flight level meteorological observations made at 15 minute intervals to support the VORCORE project
http://www.lmd.polytechnique.fr/VORCORE/McMurdo.htm

The average balloon duration was 58.5 days with the longest duration of 109 days. The scientific results of VORCORE include studies of gravity wave activity, the character of the dispersion regime inside the polar vortex, insight into ozone depletion rates in the spring and the accuracy of analyses over this region. The papers resulting from VORCORE and related preparatory efforts include Vial et al. 2001; Hertzog and Vial 2001; Hertzog et al. 2002; Hertzog et al. 2007). In terms of logistics, CONCORDIASI builds upon the success of the previous NSF-CNES collaboration in deploying from McMurdo Station.

The potential of such balloons as research platforms is improving rapidly due to a revolution in computational hardware, global positioning and communication coupled with the development of low-power, miniature sensors. For CONCORDIASI, the scientific capabilities of these balloons as sensing platforms will be dramatically improved over VORCORE. A change in the communication system used by CNES from low-band width ARGOS sensors to IRIDIUM will improve the flight level observations of meteorological parameters from 15 to 1 min intervals allowing dynamical features such as gravity waves to be directly resolved, especially considering the balloons to a first approximate are moving with the horizontal winds. Five of the CONCORDIASI flights will be dedicated to chemistry and microphysical missions designed for Four-month durations. They will be launched during August and September 2008. For these five flights, the Laboratoire de Meteorologie Dynamique at the Ecole Polytechnique in France is currently developing a small lightweight sensor able to measure ozone. In addition, four particle counters designed for the study of polar stratospheric clouds developed by Terry Deshler’s group at the University of Wyoming will be flown on these five flights.

Small backscatter/depolarization lidars will be flown as well during CONCORDIASI and will give further information on the particle sizes and phases of the polar stratospheric cloud layers and upper-tropospheric clouds below the balloons. Those balloon-borne lidar observations will complement the spaceborne lidar observations performed onboard the Calipso A-train satellite, as they will provide higher resolution and continuous sampling of the polar vortex. The lidars are under development by Guido DiDonfrancesco and Francesco Cairo at the Italian laboratory CNR.

While stratospheric balloons flown by CNES and other groups have carried remote sensing, often to test future satellite sensing systems, the addition of a micropulse lidar on long duration flights is a powerful research tool extending the scientific utility of these balloons beyond the realm of just flight level measurements. Jennifer Haase of Purdue University is planning to upgrade the CNES balloons to carry high accuracy, dual frequency GPS receivers. Such receivers have the potential to produce GPS radio occultation profiles of refractivity. GPS radio occultation was first developed for receivers on low earth orbiting satellites, which would record signals from higher orbit GPS satellites as they set behind the Earth’s limb (Kursinksi et al., 1997). As the line of sight of the GPS signal passes successively deeper into the atmosphere, the signal path is refracted (bent and delayed) as a function of atmospheric density. The refraction is measured by the Doppler shift of the carrier frequency of the GPS signal providing information on the pressure, temperature and humidity structure of the atmosphere. The refractivity is of great value for research and for assimilation in the numerical models. Jennifer Haase and colleagues are in the process of developing a similar sounding system for the new National Science Foundation HIAPER Research Aircraft (Garrison et al., 2005). The best-case scenario would provide on average about 2 occultation profiles per hour for each balloon in a region that normally has over order 20 soundings per day above 60 degrees southern latitude. In contrast to the quasi real-time continuous nature of the GPS occultation, the soundings are typically at 0000 and 1200 UTC. The coincidence in the timing between GPS occultation profiling with the release of dropsondes would be an excellent means for validating the system for each flight.

CONCORDIASI will further extend the concept of moving beyond a flight level research platform by launching up to 15 balloons in September and October 2008 that carry gondolas called driftsonde, each contain up to 60 GPS dropsondes. These dropsondes can be released upon command from the balloon flight level falling slowly to the surface producing a high-resolution vertical profile of temperature, humidity, winds and pressure. In the past, these GPS dropsondes have been extensively deployed from aircraft for research and operational prediction (Franklin et al. 1999). The dropsonde has been singled out as one of the most significant contributions to the atmospheric sciences (BASC 2006). The driftsonde has been successfully flown on CNES stratospheric balloons during the 2006 hurricane season for flights with durations over Africa and the tropical Atlantic (Parsons et al. 2007) during the AMMA field campaign

http://www.amma-international.org/rubrique.php3?id_rubrique=1

Observations over Antarctica, including the Concordiasi ones can be seen from the site

http://www.cnrm.meteo.fr/concordiasi-dataset/

In addition to the observational component, CONCORDIASI has an extensive modeling and data assimilation effort. Meteo France will conduct global and high-resolution simulations will be performed. The French global model ARPEGE (eg Fourrié et al., 2006), developed in collaboration with ECMWF, uses Four-Dimensional Variational Assimilation, and will be adapted to have a better spatial resolution over Antarctica. The French limited area model, MesoNH, or its numerically efficient version AROME, will be nested, with a grid as fine as 2.5km. Simulations will also be conducted with the Antarctic Mesoscale Prediction System (AMPS) project (Powers 2003), run jointly by NCAR and the Ohio State University. AMPS was designed to provide numerical weather prediction (NWP) support to the US Antarctic Program (USAP) and a host of foreign nations. AMPS employs the Weather Research and Forecasting (WRF) model (Skamarock et al. 2005) with a nested grids, the coarse resolution is 60-km telescoping down to an inner mesh of 2.2 km. We will also conduct simulations with a nested regional climate model under development at NCAR and collaborate with other investigators led by Andrew Gettelman at NCAR who will be undertaking climate model simulations using the NCAR Community Atmospheric Model (Collins et al. 20006). The high resolution models will prove valuable in resolving many processes critical to the climate system over the region such as vertically propagating gravity waves that influence stratospheric clouds, orographically induced circulations and mesoscale vortices that result in intense precipitation events, while the global modeling will allow investigations of the teleconnections between Antarctica and lower latitudes. The regional climate model will be the link between these two systems supporting efforts to understand how to better parameterize critical processes in the coarse grid models. The AMPS and French modeling effort will also have a strong data assimilation effort aimed at improving the analysis and reanalysis produced over Antarctica. The validation at fine scale will be performed in particular by Christophe Genthon from Laboratoire de Glaciologie et Géophysique de l’Environnement, LGGE, (Genthon et al., 2005). The impact of these improvements to the simulations and forecasts of the ozone profile will also be assessed using a chemical-transport model.

The Scientific Goals of CONCORDIASI

a) Development and testing of innovative technologies to expand the suite of polar measurements

As discussed earlier in this document, the harsh conditions (cold temperatures, high wind, blowing snow and precipitation), high elevation and remote location of continent make operational and research measurements very difficult and, in some regards, the continent is relatively data sparse. Accurate measurements of surface, meteorological parameters and chemical composition of the continent are important for documenting and understanding the key mechanisms involved in global change that impact changes to the environment, biology and human activities in polar regions. Clearly, satellite data sets are critical for Antarctica, but even the use of these systems also depends, in part, on using in-situ and local remote sensing measurements to evaluate and calibrate the data from these systems. Such observations are also critical to evaluate techniques for satellite data assimilation. Expansion of the routine surface sensing network over Antarctica is one method to accomplish the need for multidisciplinary remote and in-situ observations over the continent, but such observations are costly and it is difficult to accurately measure processes in the polar stratosphere from the surface. Research aircraft is another observational strategy, yet this approach also presents difficulties. For example, even NSF’s new HIAPER aircraft can not reach the altitudes nor have the upward looking instrumentation needed for such topics as the interaction stratospheric ozone and polar stratospheric clouds. Aviation is also expensive and polluting. The remote location is also a challenge for most airborne platforms requiring either very long duration flights (e.g., HIAPER) and/or basing in Antarctica.

The international measurement capabilities of the constellation of CNES balloons with the flight level meteorological, microphysical and chemical observations, remote sensing lidar and GPS occultation and the deployment of dropsondes is a new, and we believe extremely valuable, research platform for polar research. Launches in the winter stratospheric vortex over Antarctica is extremely well suited for such observations as demonstrated by the VORCORE effort trajectories. The intellectual merit of these developments are demanding due to the nature of long duration, stratospheric ballooning. In particular, instruments, data systems and communication capabilities on a stratospheric balloon must be low power, able to withstand extreme cold as well as diurnal temperature variations with intense solar heating, work at very low pressures, light weight and low cost and non-polluting since the systems are typically not recovered. Due to these challenges, the development phases of these proposals include a stringent testing phase with the last set of testing flights to take place over the Pacific from Brazil in January and February 2008.

b) Ozone depletion, NAT clouds and stratospheric dynamics

The southern-hemisphere polar stratosphere is the place where the largest ozone depletion takes place. Despite the implementing of international protocols that ban or limit the emission of ozone-depleting species (chlorine and bromine compounds), the recent 2006 Ozone Assessment emphasizes that the Antarctic ozone hole will remain as deep as it is now for at least 15 years (WMO, 2006). The Assessment also stresses that most of the interannual ozone-hole variability is caused by dynamical factors such as the activity of stratospheric waves that significantly modulate the background temperatures and thus the potential for Polar Stratospheric Cloud (PSC) formation. Long-duration balloons deployed during CONCORDIASI will carry instruments aimed at addressing those stratospheric issues, and in particular some of the links between stratospheric chemistry, dynamics and microphysics.

It has been shown during VORCORE (which took place in McMurdo in September and October 2005) and previous superpressure balloon flights that mesoscale gravity waves can be fully characterized from in-situ observations on-board superpressure balloons (Hertzog and Vial, 2001). Figure 4 for instance represents the momentum flux carried by gravity waves estimated from the VORCORE flights. The maps emphasize the strong wave activity observed above the Antarctic Peninsula. A recent study suggest that those small-scale orographic waves can play a key role in triggering NAT nucleation in Antarctica and thus be of primary importance for ozone depletion (Höpfner et al., 2006). During CONCORDIASI, the in-situ meteorological observations will be performed more frequently than during VORCORE (1 obs/min. instead of 1 obs/15 min.) and will therefore enables us to obtain further information on high-frequency gravity waves that are very likely generated above mountains as well. In comparison with the VORCORE dataset, the accuracy of pressure and temperature observations will also be improved. This will permit to improve our assessment of gravity-wave momentum fluxes, and thus to better characterize the role of those waves in driving the middle-atmosphere global-scale Brewer-Dobson circulation.

In conjunction with these meteorological observations, LMD is currently developing a small lightweight sensor able to perform ozone observations during long-duration flights. This ozone sensor will be hosted on 5 CONCORDIASI flights. The fact that superpressure balloons behave almost as Lagrangian tracers will enable us to directly derive ozone-loss rates along theirtrajectories. There are actually still uncertainties on the rates of catalytic reactions involved in ozone depletion: laboratory estimates of those rates are generally too small to explain the observed ozone loss (Frieler et al. 2006, Tripathi et al., 2006). Observations of temperature, position (and thus solar zenith angle) every minute, and of ozone concentration typically every 30 minutes during the CONCORDIASI flights will provide strong constraints on the catalytic rates. They will also help to determine if there are regions in the Antarctic polar vortex where ozone depletion is more likely to occur (for instance in the lee of the Antarctic Peninsula). The impact of these improvements on the simulations and forecasts of the ozone profile will also be assessed using a French chemical-transport model.

Those meteorological and chemical observations will be furthermore complemented by microphysical measurements. First, four particle counters designed for PSC study and developed by Terry Deshler’s group at the University of Wyoming will be flown on the same flights than the ozone sensors (T. Deshler already submitted a proposal to NSF for the funding of his instruments). It will thus be possible to get direct Lagrangian information on the link between particle nucleation (especially NAT nucleation) and temperature history of air parcels, and in particular to determine the relative importance of synoptic-scale and meso-scale (gravity wave) cooling in the formation of PSC, which is still uncertain (WMO, 2006). Second, two small backscatter/depolarization lidars will be flown as well during CONCORDIASI and give further remote information on the particle sizes and phases of the PSC layer below the balloons. Those balloon-borne lidar observations will complement the spaceborne lidar observations performed onboard the Calipso A-train satellite, as they will provide higher resolution and continuous sampling of the polar vortex.

c. Data assimilation

Observations and analyses of the atmosphere are essential for scientific studies at various time scales. Analyses can be used as initial conditions for real-time forecasts which can then be used to understand and predict polar meteorology, in particular in support of operations in the Polar regions. For climate research, re-analyses are very powerful tools to document past atmospheric conditions. The construction of analyses/reanalyses relies on numerical models and the fields produced by these models can be used for a variety of purposes including deriving parameters in poorly sampled areas, defining the relationship between variables and their behavior in space and time and to detect errors in the observations. The general goal of data assimilation, particularly for studies related to the climate record, is thus to incorporate as much data as possible to avoid drawing conclusions that are model dependent. Over Antarctica, the scarcity of conventional data coverage needs to be compensated by satellite data particularly those in polar orbit. Thus studies have shown that satellite data have a much larger impact in Antarctica than in other better-observed areas such as the Arctic for instance (e.g. Andersson, 2006). The satellite data impact brought successes in data assimilation. However, challenges remain and the usage of satellite observations still needs to be optimized (eg Fourrié and Rabier, 2004). This is particularly true for the new generation of advanced hyperspectral sounders, such as AIRS on Aqua, IASI on MetOp and the future CrIS on NPOESS.

The primary contribution of Meteo France to the IPY is to improve the assimilation of IASI data. Florence Rabier is the PI of that effort. IASI spectra in the infra-red possess information about temperature and humidity profiles, as well as on trace gases. The signal over polar regions is weaker than in other areas, due to the very cold atmosphere. However, absorption lines are clearly visible and signal can be extracted from these data. Key impediments to a successful assimilation of advanced sounders over these areas are cloud detection and surface contribution to the measured emissivity. As an example, clouds over very cold surfaces in a stable atmosphere will appear warmer in infrared data compared to the underlying surface. This is the opposite signal expected in most cloud detection schemes. Furthermore, polar stratospheric clouds are difficult to detect and can alias the temperature signal. Another issue is the significant variability of the polar surface (particularly in terms of temperature and microwave surface emissivity away from the more homogeneous conditions over the high plateau), since channels designed to provide temperature information in the mid-troposphere still have a 10% sensitivity to the surface. Thus errors in modelling the surface emission in these channels can be harmful for retrieving the useful atmospheric signal. As a consequence of both these issues, usage of low-peaking channels is very limited over the poles, and should be enhanced. In-situ measurements can provide ground truth to validate our assumptions/methods, in particular over inland Antarctica. This will be performed in collaboration with NCAR (Dale Barker).
Barker proposes data assimilation experiments using the WRF-Var system (Barker et al. 2004), used currently in the real-time AMPS. Firstly, we propose a series of observation impact studies for the additional dropsonde, rawinsonde, etc., in-situ data. Tests will assess both the subjective analysis response to the data for high-impact precipitation events, and the overall impact of the data through extended period trials. This work will complement NSF and NASA studies underway to assess the impact of COSMIC and AIRS data in AMPS. Secondly, we propose to assess the data quality from the Infrared Atmospheric Sounding Interferometer (IASI) during the chosen test period through the monitoring of IASI observation minus forecast differences. This work will be performed in collaboration with Meteo-France (Florence Rabier) and will make use of WRF-Var’s ability to process radiance data via EUMETSAT’s RTTOVS radiative transfer model (Liu and Barker, 2006). Thirdly, the majority of observation impact studies will be performed with WRF-Var in mode. However, selected case-studies will apply WRF-Var in 4D-Var mode (Huang et al. 2006). Additionally, WRF adjoint sensitivity studies will assess the dependence of Antarctic forecast skill on components of the analysis and the observation network (e.g., Langland and Gelaro 2004). This will be the first application of 4D-Var and its adjoint in limited-area polar NWP. When the soundings from the balloon-borne occultations are available, we will test the impact of such data sets on the accuracy of analyses and reanalyses over the region following the work of Wee et al. (2007) that utilized the assimilation of satellite occultation measurements. Finally, through collaborations with Andrew Tangborn (UMBC - see attached letter) we can compare these results with the Lagrangian and Ensemble Transform Kalman Filter techniques employed by the groups at the Universities of Maryland.

5. Predictability studies: Precipitation and Impacts on Lower Latitudes

Obtaining improvements in predictions of high impact weather over polar regions and understanding the predictability of such events is a goal of the French IPY effort. It has been shown that adding "targeted" in-situ data based on predictability information brings valuable improvement to the forecast at midlatitudes (e.g., the Winter Storm Reconnaissance Program). The general concept is to take measurements in regions where increases in accuracy of the analyses (e.g., the initial conditions) will result in improvements in simulations based on these initial conditions. For a review of this topic, see Langland (2006). A scientific question is whether this approach holds for the polar regions, especially given that the approach seems to have greater improvements in data sparse initial conditions. The driftsonde observations are ideally suited to experiment in that context. It is planned to gather information about predictability from the TIGGE (THORPEX Interactive Global Grand Ensemble) in collaboration with ECMWF and to deploy observations in the sensitive areas as well as quite uniformly. The sensitive areas will be defined relative to precipitation events. This will give us the opportunity to perform data assimilation experiments with various observation scenarios. It will then be possible to evaluate the potential of the driftsondes to improve the NWP system over the polar areas, and possibly more generally over the Southern Hemisphere. Impact at lower latitudes will be performed in particular in collaboration with BMRC (Australia).

We propose to use the data from the proposed CONCORDIASI balloon effort for two different purposes in the context of AMPS. The areas of data application would be model verification and polar physics development. The observations from the driftsondes would first provide a valuable dataset for model analysis and verification. Given the paucity of in-situ measurements of the troposphere across the Antarctic continent, the driftsonde measurements would be a unique database for the verification of the AMPS forecasts over the Ice during the study period. GPS balloon-borne occultations produced through CONCORDIASI would augment the dataset for this effort. That the driftsondes and occultations would offer continuous, asynoptic observations would support an assessment of the model’s actual forecasts not possible with the traditional radiosonde network, very limited in both time and space over Antarctica. Thus, the observations would first be used to verify AMPS performance for the period of the campaign. Second, the dataset (both driftsonde and GPS occultation) would be used in retrospective re-simulations conducted for the development of Polar WRF (Bromwich and Hines 2006). This is a version of the model which contains modifications to better capture features unique to the high latitudes, such as extensive ice sheets and sea ice. The development of Polar WRF is an ongoing effort, and would greatly benefit from application of the Concordiasi dataset for analyses of polar modifications and subsequent further tuning. The effort here would involve simulations with and without the modifications and successive validations against the database as adjustments to the polar code are made iteratively. The advancements into the understanding of the model physics will prove useful to climate modeling efforts at NCAR. We intend to use the newly developed regional climate model as a bridge between the high resolution AMPS and French MesoNH models. A particular area of focus will be to examine the effects of key processes critical to global change highlighted by CONCORDIASI measurements (e.g., polar vortex, gravity waves, circulations in the lower stratosphere). An area of focus that will be accomplished in part through collaboration will be to understand the current uncertainties in the prediction of precipitation over the continent and attempt to improve these predictions in weather and climate models. Our tools in this study include i) improving model physics and the sensitivity of precipitation processes to model physics, ii) upscaling the results of the high resolution models to global models where processes are more likely to be resolved, iii) advancing knowledge of how biases and other errors in the analysis and reanalyses result in errors in the prediction of precipitation.

A critical component of our work to understand and improve predictions of the precipitation accumulation over Antarctica (including the work done in Concordia by LGGE) will be a collaboration with the MSPICE program led by Andrew Gettelman of NCAR. The MSPICE project will use detailed in-situ observations of humidity over the Antarctic to understand ice formation and nucleation processes. Their work will also couple this information with models to estimate the net humidity convergence and thus the deposition of mass onto the Antarctic ice sheet, and extend the method for the Greenland ice sheet. The results will be a better understanding of ice formation generally, and an independent estimate of the mass accumulation onto the Antarctic and Greenland ice sheet, a major component of the ice sheet mass balance. The CONCORDIASI dropsonde profiles of temperature and humidity will be valuable for helping evaluate and constrain satellite estimates of humidity, particularly associated with AIRS. MSPICE and CONCORDIASI will work together to analyze and use the results from this multi-scale model approach to improve the treatment of processes related to the precipitation in the NCAR climate model.