DERIVING ATMOSPHERIC MOTION VECTORS FROM AIRS MOISTURE RETRIEVAL DATA David Santek 1, Sharon Nebuda 1, Christopher Velden 1, Jeff Key 2, Dave Stettner 1 1 Cooperative Institute for Meteorological Satellite Studies, University of Wisconsin, Madison, WI USA 2 NOAA/National Environmental Satellite, Data, and Information Service, Madison, WI USA Abstract Over the last 10 years, polar winds from the Moderate Resolution Imaging Spectroradiometer (MODIS) imagery have been generated by NOAA and the Cooperative Institute for Meteorological Satellite Studies (CIMSS). These datasets are a NOAA/NESDIS operational satellite product that is used at more than 10 major Numerical Weather Prediction (NWP) centers worldwide. The MODIS polar winds product is composed of both infrared window and water vapor tracked features. The water vapor atmospheric motion vectors (AMV) provide a more complete geographic distribution than the infrared window since both cloud and clear-sky features can be tracked in the water vapor images. However, polar orbiting satellites over the next decade (Metop and Suomi National Polar-orbiting Partnership (NPP)) will not have a water vapor channel as part of the imaging instrument, which will result in only infrared window AMVs. This potential gap in clear-sky AMVs provides motivation to investigate the use of Atmospheric Infrared Sounder (AIRS) moisture retrievals from consecutive overlapping Aqua satellite polar passes to extract atmospheric motion from clear and above cloud regions on constant pressure surfaces. Since the AIRS retrieval algorithm results in many tropospheric levels of moisture fields, this method also has the potential to provide vertical wind profiles, as opposed to the current MODIS-derived single-level AMVs. INTRODUCTION The study and generation of polar winds from the Moderate Resolution Imaging Spectroradiometer (MODIS) imagery was pioneered at the University of Wisconsin by NOAA and the Cooperative Institute for Meteorological Satellite Studies (CIMSS) in the early 2000s (Key et al. 2003). The MODIS polar winds product is composed of both infrared window (IR-W) and water vapor (WV) tracked features, resulting in atmospheric motion vectors (AMVs). The WV AMVs are only attainable at midand upper- tropospheric levels due to the MODIS WV atmospheric contribution function, while IR-W images also provide cloud tracers for vectors at lower levels. However, the WV AMVs yield a better spatial distribution than the IR-W since both cloud and clear-sky features can be tracked in the WV images. The next generation polar satellite era began last year with the launch of the Suomi National Polarorbiting Partnership (NPP). Unlike MODIS on Terra and Aqua, there is currently no WV channel on NPP s Visible/Infrared Imager/Radiometer Suite (VIIRS) instrument, potentially resulting in a data gap with only IR-W derived AMVs possible. This scenario presents itself as an opportunity to investigate using Single Field of View (SFOV) Atmospheric Infrared Sounder (AIRS) moisture retrievals from consecutive overlapping polar passes to extract atmospheric motion from clear-sky regions on constant (and known) pressure surfaces; i.e., estimating winds in retrieval space rather than radiance space. These winds will be combined with the MODIS imager AMVs, resulting in a blended polar winds product. The goal of the project is to determine to what extent AIRS-derived AMVs can provide useful wind information. Even though the spatial resolution of the images is much reduced compared to MODIS (16 km vs. 2 km), there are several potential advantages:
1) Whereas the MODIS winds are at a single level, the AIRS winds will provide a 3-dimensional winds dataset. 2) The AIRS retrieval algorithm results in moisture features at constant pressure levels, thereby removing issues with the MODIS AMV height determination. 3) Clear sky (and above cloud) wind information can be obtained, which will be important in the post-modis era of instruments with no water vapor imager channel. Finally, NWP experiments will be run with the blended product to determine the overall impact on numerical forecasts, and the relative contributions of each data type (MODIS vs. AIRS). AIRS RETRIEVAL PRODUCT The AIRS Standard Retrieval Product provides profiles of retrieved temperature, water vapor, and ozone. The profile vertical resolution is 28 levels between 1000 and.02 hpa (Susskind et al. 2003). This standard product is generated from 3x3 FOVs of AIRS radiances (Goldberg et al. 2003), which results in a horizontal resolution of 40 km. This is much too coarse for tracking features from successive orbits as a 1-pixel displacement corresponds to 6.7 m s -1. Weisz et al. (2007) at CIMSS have developed a single FOV (SFOV) retrieval algorithm that retains the native horizontal resolution at 13.5 km/pixel and results in temperature and humidity retrieved at 101 pressure levels from 0.005 to 1100 hpa. Comparisons of these profiles to collocated radiosondes observations and model analyses are similar to that of the Standard Retrieval product (Weisz et al. 2007). AIRS AMVS The AIRS retrieval algorithm produces vertical profiles of specific humidity and ozone concentration on constant pressure surfaces. Next, images of these retrieved quantities on each pressure surface are generated. For example, the specific humidity at 359 hpa for 7 January 2005 at 1041 UTC is depicted in Figure 1. This image is overlaid with wind vectors derived from the motion of these moisture features. Figure 1: AIRS Retrieval Images at 359 hpa specific humidity SFOV AIRS retrievals. Remapped composites at 16 km resolution. Date: 7 January 2005 at 1041 UTC. Because of the reduced spatial resolution and narrower swath, as compared to MODIS, there are many fewer wind vectors at each image time. However, when viewed over the span of a day, there is
good spatial distribution (Figure 2). Also, at many locations there is a vertical profile of wind information, which is not evident in Figure 2. See Figure 5 for an example vertical distribution. Figure 2: All derived winds from 5 January 2011. Color coded by level: 700-600 hpa (red); 550-450 hpa (green); 400-300 hpa (blue); 150 hpa ozone (gray). ASSIMILATION Since there are very few conventional observations (e.g., rawinsondes) over the polar regions, a comparison with the GSI background was done to measure the quality of the winds. This first comparison is using all winds without any quality control. A time period from 01 14 January 2011 was chosen over the northern hemisphere. 29 levels of winds were derived: 12 ozone and 17 moisture levels (away from tropopause). The ozone levels are in the range 103-201 hpa; the moisture from 359 661 hpa. A 2010 version of the Gridpoint Statistical Interpolation (GSI) 3D-Var system was used for the assimilation. Figure 3 is a histogram of the observation minus background (OMB) for all, moisture, and ozonederived winds. Although the standard deviation is high at about 10 m s -1, it is encouraging that the bias is very nearly zero.
Figure 3: Histogram of Observation minus Background (OMB) for all (gray), moisture (green), and ozone (blue) winds for the two-week period 01-14 January 2011. The histogram in Figure 4 is the observation minus analysis of the winds assimilated. The standard deviation has been reduced to less than 7 m s -1, however, the bias is about 3 m s -1. The large bias is thought to be the result of using all the winds without any quality control. Figure 4: Histogram of Observation minus Analysis (OMA) for all (gray), moisture (green), and ozone (blue) winds for the two-week period 01-14 January 2011. The assimilated winds are shown as a vertical profile for one assimilation cycle on 6 January 2011 at 1200 UTC (Fig. 5). This shows a good vertical and longitudinal distribution, but it also depicts some cases of large differences from the analysis (gray vectors), which again is the result of no quality control at this time. When the blended MODIS and AIRS winds product is generated, the MODIS winds will be used in the AIRS quality control step to filter out erroneous vectors.
Figure 5: Vertical distribution of AIRS retrieval winds used in the assimilation over the north pole region. All derived winds from 6 January 2011 at 1200 UTC. Colors denote distance from pole: blue (far) to red (close). Gray is the analysis. STATUS The above results are very preliminary as we continue to refine the procedure. Noisy retrievals have required smoothing of the 2D moisture images; we are transitioning from a low-pass to a median filter. Also, we continue to fine tune settings in the wind retrieval algorithm to increase the number of quality controlled vectors. REFERENCES Goldberg, M. D., Y. Qu, L. M. McMillin, W. Wolf, L. Zhou, M. Divakarla, 2003. AIRS near-real-time products and algorithms in support of operational numerical weather prediction. IEEE Trans. Geosci. Remote Sensing, 41, pp. 379-389. Key, J. R., D. Santek, C. S. Velden, N. Bormann, J.-N.l Thépaut, L. P. Riishojgaard, Y. Zhu, and W. P. Menzel, 2003. Cloud-Drift and Water Vapor Winds in the Polar Regions from MODIS. IEEE Trans. Geosci. Remote Sens., 41, pp. 482-492. Susskind, J., C. D. Barnet, J. M. Blaisdell, 2003. Retrieval of atmospheric and surface parameters from AIRS/AMSU/HSB data in the presence of clouds. IEEE Trans. Geosci. Remote Sensing, 42, pp. 390-409. Weisz, E., H.-L. Huang, J. Li, E. Borbas, K. Baggett, P. Thapliyal, and G. Li, 2007. International MODIS and AIRS Processing Package: AIRS products and applications. J. Appl. Remote Sens., 1, pp. 1-23.
ACKNOWLEDEMENTS This work is supported under NASA Grant NNX11AE97G.