Protecting Our Nation with Research That Makes a Difference Contents

VOLUME 29, SPRING 2008

Protecting Our Nation with Research that Makes a Difference contents

Geo-Targeted Alerting System (GTAS) Members of the Advisory Board Page 1 Peter K. Dorhout, Colorado State University Vice Provost for Graduate Studies Data Assimilation: Extracting Bill Farland (Chairperson), Maximum Information from Colorado State University Vice President for the GOES-R Data Research Richard Johnson, Page 4 Colorado State University Atmospheric Science Department Head CIRA’s Multisensor Blended Total “Sandy” E. MacDonald, Precipitable Water Products Deputy Assistant Administrator for Laboratories and Cooperative Institutes/Director, ESRL Serve Forecaster Needs Al Powell, Page 8 Director, NOAA/NESDIS/STAR Thomas H. Vonder Haar, Founding Director of CIRA and University Regional Impacts of Oil and Gas Distinguished Professor, Colorado State University Development in the Department of Atmospheric Science (ex officio) Western United States Sandra Woods, Colorado State University Dean of Page 12 Engineering Members of the Advisory Council Environmental Education and Public Hal C. Cochrane, Outreach Professor, Colorado State Department of Economics Page 20 Mark DeMaria, Team Leader, NOAA/NESDIS/RAMM Ingrid Guch, Cloudsat DPC Hardware Chief, NOAA/NESDIS/CoRP Page 24 Steven Koch, Director, Global Systems Division/ESRL Photos above, from top: Sonia M. Kreidenweis-Dandy Figure 1. Surface pressure ESRL’s New Global Model and Its Professor, Colorado State Department of forecast difference (hPa) Atmospheric Science between the experiments with Implementation and without data assimilation. Page 26 Thomas H. Vonder Haar (Chairperson) Figure 2. The blended TPW Founding Director of CIRA and University AMSU and SSM/I product Distinguished Professor, Colorado State University augmented with GPS data over Department of Atmospheric Science the CONUS on March 4, 2008. Figure 3. CIRA’s Director’s Magazine Staff Conference Room after the remodel. Figure 4. Wind speed Mary McInnis-Efaw, Editor (meter/sec.) at 500mb and Laura Leinen, Assistant Editor geopotential height at 500mb. Design Colorado State University, Communications and About the cover: Testing the Creative Services Geo-Targeted Alerting System Cooperative Institute for Research (GTAS) with a simulated toxic in the Atmosphere Technical Committee release over San Francisco. Photo courtesy of Will von Cliff Matsumoto, Donald Reinke, Renate Brummer, Dauster of ESRL/GSD. www.cira.colostate.edu Marco Rodriguez GTAS Geo-Targeted Alerting System (GTAS) Herb Grote Since the events of September 11, 2001 the atmosphere. AWIPS in New York City, considerable focus has been is currently being rede- placed on protecting the public from terrorist signed to use a Service attacks and other man-made disasters that could Oriented Architecture jeopardize the lives of a large number of people. (SOA) and is expected The release of biological, chemical, or radioac- to be ready for initial tive material into the atmosphere in a heavily deployment in fiscal year populated area could potentially kill thousands of 2010. This presents an people. The GTAS project is designed to develop opportunity to incor- an advanced warning system that can rapidly and porate essential new precisely warn and provide safety information capabilities to support and instructions to the public in case of a life- GTAS requirements. threatening toxic release. The geo-targeted A key element of NOAA’s current mission is concept came from to alert the public of potentially life-threatening exploratory work using weather events, such as blizzards, tornadoes, and reverse-911 notifica- flash-floods. The National Weather Service has tions to provide highly the infrastructure in place to forecast the weather localized notifications Figure 1. A sample GTAS screen and to disseminate the forecasts, watches, and of a hazardous weather event such as a tornado to showing some of the MADIS warnings to the public in an effective and timely affected individuals. To make geo-targeting fea- surface observations for the manner. Weather offices exist in all fifty states and sible requires a telephone database that is current, Washington, D.C. area. often are located near major population centers. complete, and contains very specific geographic Over the years, forecasters in these offices have location information in addition to just a street established local contacts, developed procedures, name and address. The initial vendor to support and gained significant experience in supporting this exploratory work was Intrado, Inc. located local governments during hazardous weather in Longmont, Colorado. The FX Collaboration events. That same infrastructure can be leveraged (FXC) system developed by the Global Systems to deal with terrorist attacks and other man-made Division at NOAA Earth System Research Lab was disasters. used to graphically generate a warning area and The NOAA infrastructure includes advanced electronically transmit the geographic coordi- new technologies that can generate high reso- nates in latitude and longitude to the vendor. The lution weather data for use by Department of vendor then accessed the telephone database and Homeland Security (DHS) operations. Some of determined all of the phone numbers within the the NOAA technologies designed to be used by geographic area. After operator confirmation, GTAS are the Advanced Weather Information the vendor’s system dialed the designated phone Processing System (AWIPS), the Meteorological numbers as quickly as possible according to a Assimilation and Data Ingest System (MADIS), simple rule, e.g., from north to south. The pur- the Graphical Forecast Editor (GFE), the Hybrid pose of the test was to determine the time it took Single Particle Lagrangian Integrated Trajectory to make the first phone call and the rate at which (HySPLIT) model, and the 4-km Nonhydrostatic calls could be made thereafter. Obviously, if the Mesoscale Model (NMM). The AWIPS system is vendor was unable to make the first call before the in all of the NWS forecast offices and provides event entered the area or unable to call all of the forecasters with a powerful tool to view and numbers before the event traversed the area, the understand the past, present, and future state of system was deemed of limited use.

1 GTAS

DHS’ FEMA quickly became interested in weather forecast models are available to initial- developing a public warning concept for poten- ize the dispersion model. Upon completion, the tial toxic release in the National Capital Region client is notified and the dispersion plume can be based on our reverse-911 geo-targeting work. viewed as a vector graphic or image. The prototype system is based on FXC, which Another feature added to GTAS was the has many of the features of AWIPS. FXC is well ability to send the warning area polygon to suited for developing this prototype since it can multiple reverse-911 vendors. GTAS needed the be modified easily, unlike AWIPS which is under flexibility to communicate with different vendors software configuration management by the NWS. according to their unique network and security It also has many of the features required by GTAS, requirements. The same message, in the form such as weather data display and interaction, of a simple custom XML document, is sent to real-time collaboration, graphical annotation, and all of the vendors. In the final version, the XML secure transmission of data to remote servers. The message would comply with the OASIS Standard FXC capability allows participants at physically Common Alerting Protocol (CAP). The ability different locations to view identical weather data to notify several vendors can be used in areas display and, through interactive tools, help in the where a single vendor does not have a complete discussion of the forecast. Also, since FXC is writ- telephone database for the area, or to selectively ten in Java, the same programming language used determine which of several vendors provides the by AWIPS II, it will make it easier to transfer code best performance. GTAS supports encrypted https to the new AWIPS system. Some portions of the with security certificates and keys. FXC graphic annotation code have already been In order to accurately initialize the dispersion ported to AWIPS II. model and draw an accurate warning polygon, the FXC needed some additional features in map background information must be accurate order to be used as a functional prototype for and of high spatial resolution. The initial maps GTAS. The first and most significant feature to obtained from AWIPS had insufficient detail on be added was the ability to run the HySPLIT the city scale. In order to meet the GTAS require- dispersion model. The HySPLIT dispersion model ments, high resolution (ESRI) shapefiles were Figure 2. View of the HySPLIT calculates the dispersion assuming either puff or downloaded and mapped to the Washington, dispersion model dialog box superimposed on the dispersion particle dispersion. The model’s default configu- D.C. area. A utility to read shapefiles was partially plume generated by the model. ration assumes a puff dispersion horizontally implemented using the open source Java GIS and particle dispersion toolkit GeoTools. Additional work is required to vertically. The model is obtain greater benefit from the available shapefile initialized on the GTAS information. client using a dialog Additional enhancements were made to deal window that allows the with firewalls and Network Address Translation user to specify various (NAT). A message repeater, unique to FXC, routes parameters such as dura- all messages to and from the FXC server processes tion of the run, maxi- through a single port. This port was opened to mum height, and the the outside and used address translation to hide release point by entering the internal network address from outside users. the latitude and the lon- An address translation table implemented by FXC gitude directly or select- can now deal with traffic through NATs. ing the location with the GTAS uses a client-server architecture with mouse cursor (see Figure dedicated T1 communications links between the 2). This information is server and clients. The dedicated lines provide then transmitted to the reliable performance and an acceptable degree of GTAS server where the security for the prototype. The server has access

2 to the full AWIPS real-time meteorological data and can serve a number of clients who are either collaborating or working independently. Dual servers are implemented to provide the necessary redundancy. For the initial prototype, the servers reside at the NOAA/ESRL Global Systems Divi- sion (GSD) computer facility. The GTAS client software can be downloaded from a local server and installed on a remote client using a simple installation script. A single remote client was installed at the FEMA office in the Washington, D.C. area and several clients were configured at GSD in Boulder. Thus far, the system has only been used to demonstrate the basic GTAS concept. FEMA staff has been able to simulate a toxic release and initialize the dispersion model. Separate tests have been conducted to draw warning polygons and send the coordinates to two reverse-911 vendors – 21st Century and Intrado, Inc. Figure 3 illustrates a dispersion plume in the Washington, D.C. area using a simulated release point and agent. For the initial prototype, the HySPLIT model is initialized using the NAM-12 grids. Animated GIF images of the dispersion plume with time have also been made available on special websites using GTAS’ capability to capture the screen and post the images to the web server. Figure 3. Example of an The GTAS project is a joint endeavor annotated plume depicting the between the National Oceanic and Atmospheric warning polygons that define Administration (NOAA) and the Department the XML message sent to the vendors. The 911 icon in the of Homeland Security (DHS). The project was upper right hand corner is started nearly 3 years ago and is now awaiting a used to transmit the polygon decision by DHS to continue with the develop- information. ment and evaluation. Acknowledgements CIRA has the technical lead for the design and implementation of the GTAS system and also performs some of the development. Other groups have made significant contributions to the project. Chris Golden (SRG) performed much of the FXC software enhancement and Xiangbao Xing (CIRES) leads the installation of the HySPLIT model. The project manager is Richard Jesuroga (NOAA).

3 DATA ASSIMILATION Data Assimilation: Extracting Maximum Information from the GOES-R Data Dusanka Zupanski, Milija Zupanski, Renate Brummer and Mark DeMaria (NOAA/NESDIS) Next Generation research is to prepare algo- prepare methodologies capable Geostationary Satellites rithms to maximize the use of answering these questions The next generation GOES of information from future before the GOES-R observa- will begin with GOES-R, which GOES-R observations, espe- tions are available. is currently scheduled to launch cially in the area of operational The answers to the ques- in the year 2015. GOES-R will severe weather forecasts. tions posed above are not The major objective of our be equipped with the Advanced Challenges of Data trivial, since the same observa- specific data assimilation Baseline Imager (ABI), an Assimilation tions could bring more or less research is to prepare imager that has significantly Data Assimilation is the information, depending on algorithms to maximize the improved spectral, spatial, and technique of blending data where and when the observa- use of information from temporal resolution relative to from very different sources in tions are taken, and on how the current GOES I-M series order to produce a new data set accurate the forecast is for future GOES-R observations, satellites. These improvements that is most consistent with the that place and that location. In especially in the area of will greatly enhance our ability atmospheric state. Advanced other words, the information operational severe weather to make mesoscale weather, data assimilation methods, from observations is dependent forecasts. climate, oceanographic, and which extract information on the number and quality of environmental observations. about the true atmospheric the observations and on the Because of the long lead- state from observations and forecast uncertainty. Unlike time required to design, build, combine it with the informa- the data assimilation methods and test new and complex satel- tion from a forecast model, currently used in operational lite equipment, preparations for have proven to be powerful weather centers (in the U.S. and GOES-R applications are well tools in improving weather worldwide), the next gen- underway. The approach for forecasts. New observations eration ensemble-based data these “Risk Reduction” activi- from GOES-R will pose new assimilation methods are dem- ties is to use data from existing challenges to the data assimila- onstrating a qualitatively new operational and experimental tion methods. A major issue capability to address the above satellites to create subsets of that needs to be addressed is challenges. Of special impor- observations that resemble to estimate if, and how much, tance is the new capability to those we will eventually receive the GOES-R observations estimate and use realistic fore- from GOES-R. These simulated are expected to improve our cast uncertainty, which depends GOES-R datasets will then be knowledge about the true on the atmospheric conditions. used for the development of atmospheric state. In order This forecast uncertainty is algorithms for new satellite to address this challenge, one typically defined in terms of a products or for the improve- needs to quantify (1) what is covariance matrix measuring ment of existing products, par- the amount of information forecast errors, the so-called ticularly for atmospheric and provided by the currently avail- “flow-dependent” forecast surface-related phenomena, able meteorological observa- error covariance. At CIRA, we using the additional GOES-R tions (e.g., conventional surface are developing and exploring resolution capabilities that will and upper-air observations) ensemble-based data assimila- be available. and (2) how much new infor- tion methods for applications The major objective of mation will be brought by the to the future GOES-R observa- our specific data assimilation GOES-R data. Our goal is to tions and to other current and

4 upcoming satellite missions Figure 1a-c. Surface pressure (e.g., CloudSat, Global Pre- forecast difference (hPa) between the experiments with cipitation Mission-GPM and and without data assimilation. Orbiting Carbon Observatory- In the experiment with data OCO). assimilation, a 6-hour forecast after data assimilation is used. Ensemble Data Assimilation In the experiment without data Ensemble-based data assimilation, a longer-term assimilation methods employ forecast is used, initialized ensembles of forecast model at the beginning of the data assimilation experiment runs to estimate flow-depen- (first data assimilation cycle). dent forecast uncertainty. Both experiments used the Using flow-dependent forecast same boundary conditions. error covariance is essential for The triangle denotes the best extracting maximum informa- position of Hurricane Katrina estimated by the NOAA/ tion from each observation National Hurricane Center because in the areas of high report. The panels cover last forecast uncertainty (where 24 hours before the landfall: ensemble members differ (a) Aug. 28, 2005 at 0000 UTC, which corresponds to 8th among each other the most), data assimilation cycle, (b) the information given by the Aug. 28, 2005 at 1200 UTC, observations is more impor- which corresponds to 10th tant. The opposite is true in the data assimilation cycle, and (c) Aug. 29, 2005 at 0000 UTC, areas where the forecast uncer- which corresponds to 12th data tainty is low (i.e., all ensemble assimilation cycle. Negative members are similar). values (blue and purple colors) Under the GOES-R indicate lower pressure, while positive values (yellow and red research project and various colors) indicate higher pressure other research projects at CIRA, in the experiment with data we are employing and further assimilation. Note that the developing an ensemble-based data assimilation is moving the cyclone towards the observed data assimilation method, location, since the triangle referred to as the Maximum is located in the blue (lower Likelihood Ensemble Filter pressure) area. (MLEF). For the GOES-R application, we are focusing on the MLEF capability to extract maximum information from the future GOES-R observations. This is accomplished by evaluating information measures of assimilated observations. Information measure called Degrees of Freedom for Signal (DFS) is often used in meteorological applications. The DFS is a number, defined for a selected set of observa-

5 DATA ASSIMILATION

tions, which counts how many in a well-tested data assimi- Does More High Quality Data pieces of independent infor- lation algorithm capable of Mean More Information? mation the observations are extracting maximum informa- We expect that the bringing into the data assimilation from the GOES-R observa- GOES-R observations will be tion system. The DFS cannot tions, before these observations of excellent quality and will be larger than the number of become available. have high spectral, spatial, and observations. In ensemble data We have recently applied temporal resolution. Can we assimilation methods, the DFS the MLEF to improve fore- automatically assume that more cast simulations of Hurricane high quality data will result in Katrina by assimilating con- improved data assimilation and ventional surface and upper-air forecast results? Classical data observations into the Weather assimilation methods, which Research and Forecasting do not employ flow-dependent (WRF) model. In the experi- (but prescribed) forecast error ment presented, the model covariance, would yield “yes” spatial resolution is 30 km as an answer. This answer, with 28 vertical levels (totaling even though desirable, may 75x70x28 grid points) and we not be correct. Modern data employ 32 ensemble members. assimilation methods, such Observations are assimilated as ensemble-based methods, every 6 hours, during 12 could provide a different consecutive data assimilation answer because of the use of cycles. Figure 1a-c shows an the flow-dependent forecast example of the impact of data error covariance. assimilation on improving the We have calculated the Figure 2. Visible satellite image forecast of Hurricane Katrina. DFS for the same case of of Hurricane Katrina on 28 Aug is also limited by the ensemble Most significant improve- Hurricane Katrina, given in 2005 at 1215 UTC. size, thus appropriate ensemble ments are obtained by placing Figure1a-c. The DFS results are size should be determined the cyclone center to the more shown in Figure 3. The model in such applications. This correct location, as indicted integration domain was divided information measure, among in the figure by the dipole into nine equally sized areas others, is used in our GOES-R (positive-negative) pattern in (sub-domains) and the DFS research. the pressure difference field, are calculated for each sub- In the first stage of this while the improvements in the domain. The sizes of the sub- research, we applied the MLEF cyclone intensity were smaller. domains are selected by taking to quantify the information These improvements are quite into account the spatial scale of content of conventional obser- substantial considering that the tropical cyclone. One can vations. In the second stage, only a relatively small number immediately notice that the which is currently underway, of observations are assimilated. highest amount of information we will examine and further For illustration, we also include is provided by the observations develop this methodology in a visible satellite image of Hur- from the central sub-domain application to satellite observa- ricane Katrina (Figure 2), valid (pink color indicates that the tions with similar character- approximately at the same time DFS is equal or larger than 14). istics as the future GOES-R as the surface pressure differ- This is exactly the sub-domain observations. We expect that ence field given in Figure1b. where the cyclone center is the second stage should result located. Note that the number of observations (the number

6 of crosses) is not the highest thus the data assimilation Defense (DoD) Center for Geo- in this sub-domain, thus the experiment will be computa- sciences/Atmospheric Research forecast uncertainty played a tionally more demanding. It (CG/AR). This research is a dominant role in defining the will be interesting to examine if team effort including, besides information content. Note also the data assimilation algorithm the authors of this article, Louie that there are more observa- will be able to correctly extract Grasso of CIRA, Sara Q. Zhang tions in the three sub-domains information from the satellite of NASA/GMAO, Arthur Y. located in the northern part of observations, while taking into Hou of NASA/GMAO, Isidora the model domain, but these account the flow-dependent Jankov of CIRA/NOAA/ESRL, observations carry less infor- forecast uncertainty. Daniel Lindsey of NOAA/ mation because the values of NESDIS, Managit Sengupta of the DFS are smaller (colored Acknowledgements CIRA, Thomas Vonder Haar in orange and green). These The research reported of CIRA, Christian D. Kum- results demonstrate the ability here was mainly funded by the merow of CSU/Department of of the MLEF to correctly take NOAA/NESDIS/GOES-R Risk Atmospheric Science and Scott into account the characteristics Reduction Program, under Denning of CSU/Department of the atmospheric flow when Grant No. NA17RJ1228. We of Atmospheric Science. The extracting information from also acknowledge research authors of this article would the available observations, thus efforts under synergistic like to thank Ray Zehr of maximizing the benefits of the research projects at CIRA, NOAA/NESDIS for preparing assimilated observations. funded by the NASA/Global the satellite images of Katrina, Additional informa- Precipitation Mission (GPM) and to Don Hillger of NOAA/ tion can be found at www. program, the NASA/North NESDIS and Laura Leinen and cira.colostate.edu/projects/ American Carbon Program Mary McInnis-Efaw of CIRA ensemble/ (NACP) and the Department of for the final editing. What is Next? The experimental results Figure 3. The DFS calculated with conventional upper-air in data assimilation cycle 10 (valid at 12 UTC Aug. 28, and surface observations were 2005), shown, in shading, for quite encouraging. The next each of the nine sub-domains step of our research is to test covering the model domain. the same methodology in Also shown are the locations of the assimilated observations application to satellite observa- (crosses) and the location of the tions. We have already selected tropical cyclone, as defined by an interesting severe weather the isolines of the magnitude of case (a very strong extra- the horizontal wind (in ms-1), obtained by the analysis at the tropical cyclone named Kyrill model level corresponding to a that occurred over Europe height of 700 m. during January 18-19, 2007) and started collecting satellite data from the Meteosat Second Generation (MSG) satellite that resemble those we will receive from the future GOES-R instrument. In this case, the WRF model will be run in finer resolution (15 km/49 levels),

7 HSEERVINGADLINE FORECASTER NEEDS How can we make the CIRA’s Multisensor Blended Total Precipitable Water Products powerful instruments onboard polar orbiting Serve Forecaster Needs John Forsythe, Stan Kidder, Andy Jones; Sheldon Kusselson (NOAA/NESDIS Satellite spacecraft show the time Analysis Branch) evolution of features which is so valuable to Forecasters today are faced with many concerns about which spacecraft or algorithm was sources of data. What they need is meteoro- used. There are many spacecraft in orbit which forecasters? logically significant data fields blended from all measure TPW and are candidates for blending. available data sources, not numerous maps of Passive microwave sensors are particularly well- observations from individual sources. Forecasters suited for measuring TPW since they can retrieve will use whatever products help them understand TPW in the presence of clouds. Clouds block the the current situation and make a better forecast, retrieval of TPW from infrared sensors such as but if the product is not easily understandable the GOES Sounder, which is unfortunate since within a few seconds it will not be used in a busy often the most interesting forecast challenges are forecast office. in cloudy areas where heavy precipitation might CIRA scientists Stan Kidder occur. But there is a catch – the only passive and Andy Jones have developed microwave sensors in space for weather measure- a technique to blend satellite ments are on polar orbiting satellites. Polar- retrievals of total precipitable orbiting data are not as friendly to forecaster use water (TPW) to make a product as geostationary data. An area such as the west which is easy to use by forecast- coast of the U.S. is observed at irregular times, ers (Kidder and Jones, 2007). so adding a regularly spaced time interval to TPW is the amount of water examine the data, such as in the loops commonly vapor in a column from the used for GOES data, is not feasible. In addition, surface of the earth to space (in in midlatitudes there may be gaps in the geo- Figure 1. Example of the six kilograms per square meter or, equivalently, in graphic area observed between one overpass and polar orbiter microwave millimeters of condensate). On a global average, the next. How can we make the powerful instru- retrieved blended TPW product from March 1, 2008. Notice the the atmosphere contains 25 mm of TPW, with a ments onboard polar orbiting spacecraft show the plumes of moisture between the range from near 0 mm (Antarctica) to extreme time evolution of features which is so valuable to tropics and midlatitudes. TPW instances of 90 mm (strong tropical cyclones). It forecasters? is not currently retrieved over is this moisture that provides the fuel for many land or in heavy precipitation. of the key elements of a forecast (clouds and Products precipitation). TPW integrated from radiosondes a) AMSU and SSM/I Blended TPW Figure 2. The blended TPW has historically been used by forecasters to predict Among the workhorse microwave instru- AMSU and SSM/I product heavy precipitation, especially in summertime ments for measuring TPW are the AMSU augmented with GPS data over flash flood situations. But radiosonde data is the CONUS on March 4, 2008. onboard the NOAA and European METOP space- Severe weather was occurring sparse compared to satellite data. craft series and the SSM/I and SSMIS onboard in the moisture plume over the The idea behind this work, funded by the the DMSP spacecraft. For the past few years, there eastern U.S. NOAA Product Systems Development and Imple- have been up to six of these spacecraft available mentation (PSDI) from an orbital height of 860 km. All of these Program, is to present instruments are very capable of measuring TPW the forecaster with the over ocean. Over land, retrieving TPW from pas- best picture of TPW sive microwave measurements is very challenging at a given time, and to due to the variable emissivity of the land surface. insulate the fore- CIRA is active in research to retrieve moisture caster from technical over land as well, but at this time only micro-

8 HEADLINE

wave retrievals of TPW over water are performed routinely. Figure 1 shows an example of the global TPW from six satellite instruments (AMSU and SSM/I). The powerful DPEAS data processing system (Jones and Vonder Haar, 2002), which was developed at CIRA, is used to manage the remapping and compositing of the data. All of the science code resides within DPEAS, allowing developers to have a common programming framework. The blended TPW product is updated every six hours and is available on amsu.cira. Figure 3. The blended colostate.edu/TPW. It is produced globally (except for weather analysis. While research on pas- microwave / GPS product tracks near the poles) on a 16 km resolution Mercator sive microwave retrievals over land should yield the high TPW associated with projection. The global product and sectors over benefits in the future, they are not yet ready for the remnants of Tropical Storm Erin from August 17 – 19, 2007. various ocean basins are available on that site. The operations. Fortunately, over the CONUS (Con- The 24 hour rainfall record for animations online capture the flow of moisture in tinental U.S.) there exists a rich dataset of very Minnesota was broken when a way not possible with other satellite products. accurate TPW measurements. You might even use 15.11” of rain fell in the 24 The familiar geostationary water vapor channel the same sensor yourself! hours ending 12 UTC August 19, 2007 in SE Minnesota. animations are a complement of this product, but The signals transmitted from the 24 GPS they only show upper level (above about 500 hPa) satellites, while intended to provide earth loca- water vapor and provide no information when tion, suffer some signal degradation due to the there are high clouds. temperature and moisture structure of Earth’s As detailed in Kidder and Jones (2007), a atmosphere. This variation in signal (called the number of corrections need to be made to the wet delay) can be used to estimate TPW above AMSU and SSM/I products to make them useful a ground-based sensor with accuracy of greater for forecasters. Even though they use essentially than 2mm, in all weather conditions. A precise, CIRA has received a wide the same spectral region to retrieve TPW, AMSU collocated measurement of atmospheric pressure range of positive feedback and SSM/I are quite different in their scanning is also required. Various government programs, geometry (cross-track vs. conical scan), and this like NSF’s Suominet (www.suominet.ucar.edu) and encouragement on the can lead to artifacts and seams where they are and NOAA’s GPS-MET network (www.gpsmet. experimental products from merged together. Forecasters do not want to see noaa.gov/jsp/index.jsp) have placed these GPS NOAA offices such as the artifacts of the sensor type in a blended product; sensors in the CONUS and adjacent regions – SPC, NWS Western Region, they just want to see the meteorology! A correc- over 400 as of the spring of 2008. John Forsythe, NHC, and HPC. tion factor is constantly updated to account for Stan Kidder, and Andy Jones have been routinely differences in the two sensors. bringing in the GPS TPW stations from NOAA/ The blended TPW product has become a GSD in Boulder since 2006. Seth Gutman of staple of the Daily Weather Discussion at CIRA NOAA/GSD has been of great assistance in (see CIRA Newsletter, Spring 2006). obtaining this data. The data are overlaid on the b) Addition of GPS and GOES Sounder over CONUS and brought together with the satel- CONUS lite microwave retrievals to form a land/ocean The global TPW product shown in Figure 1 satellite TPW product. An example is shown in has proven its value for forecasters since its incep- Figure 2. Animations of the last three days of this tion in 2003. The product is being generated product, updated every six hours, are available at: operationally at NOAA/NESDIS. But the lack of amsu.cira.colostate.edu/gpstpw. Finally, GOES information over land has hindered its usefulness sounder TPW from CIMSS at the University of Wisconsin is added when available in areas where

9 SERVING FORECASTER NEEDS

the GPS network is still sparse (for instance, parts • Tracking tropical waves which might become of Montana and Mexico). hurricanes in the Atlantic, and also recognizing This experimental product with GPS and dry air that inhibits their formation. GOES Sounder went online as a near-realtime A large injection of moisture associated with experimental website in 2006. Since then, Tropical Storm Erin occurred in August 2007. McIDAS files are being generated and served to The blended TPW field with the GPS input over NOAA/NESDIS/SAB every six hours. Sheldon CONUS allowed forecasters to track this tropical Kusselson of SAB, who recognized the value of moisture as it traveled into the upper Midwest the blended TPW product since he has forecast and caused severe flooding, as shown in Figure 3. responsibility for heavy precipitation events, has been instrumental in popularizing the product The Future with other NOAA offices and forecasters. CIRA Along with the development of the blended has received a wide range of positive feedback TPW product, an experimental TPW anomaly and encouragement on the experimental product product has been developed at CIRA. The from NOAA offices such as the SPC, NWS anomaly is the percent of the weekly normal Western Region, NHC, and HPC. TPW derived from the global NASA Water Vapor (NVAP) dataset (Randel et al. 1996) for 1988 – Forecast Applications 1999. The blended TPW product is invaluable for The progression of a large positive TPW understanding the synoptic scale flow of moisture anomaly (> 200% of normal shaded as yellow) Figure 4. TPW anomaly product in the atmosphere. Applications where the prod- is shown in Figure 4. Blue shades are more moist and GOES water vapor imagery uct shines include: than normal, and brown shades are drier than for an intense midlatitude • Atmospheric rivers striking the U.S. West Coast normal. In this case from February 2008, severe cyclone on February 16 and 17, 2008. Yellow color in the TPW and causing floods. weather occurred over Mississippi and Georgia as anomaly is > 200% of normal • Return of Gulf moisture to the central U.S. for the intense midlatitude cyclone tracked towards and is indicative of heavy clouds severe weather. the northeast. The TPW anomaly, and the com- and precipitation. panion GOES water vapor imagery, captures the location of clouds and precipitation very well, and the drying in the wake of the system is also well- articulated. Forecaster experience at NESDIS SAB indicates that the TPW anomaly field is a useful tool for forecasters to diagnose heavy precipitation events. John Forsythe and Stan Kidder at CIRA are currently exploring the utility of the TPW anomaly for forecasting. An exciting technology transfer effort is now underway to install the blended satellite/ GPS/GOES Sounder TPW product into AWIPS Operational Build 9 in 2009. Stan Kidder is leading this effort for CIRA, and Limin Zhao and John Paquette are the technical leads for NOAA/ OSDPD. Work is also underway on adding the European METOP TPW data into the blended product. As the NPOESS era approaches with the scheduled launch of NPP in the 2010 timeframe, the system needs to be agile to accept the ever- changing cast of satellite and ground-based TPW

10 sensors. The blended TPW project has served as a good example of a NOAA Cooperative Institute AMSU Advanced Microwave Sounding performing research relevant to the NOAA mis- Unit sion and implementing that research into opera- AWIPS Advanced Weather Interactive Pro- tions. It also illustrates the need for a successful cessing System technology transfer to have a recipient (NESDIS DPEAS Data Processing and Error Analysis The production code is SAB in this case) willing to make the effort to System being ported to NOAA/ try out an experimental product and provide HPC Hydrometeorological Prediction NESDIS so that the blended valuable feedback. CIRA and NESDIS are excited Center TPW product will be about future collaborations which will continue MetOp Meteorological Operational Satellite available in Operational to give NOAA forecasters the benefit of the latest NHC National Hurricane Center university research. NPOESS National Polar-orbiting Opera- Build 9 of AWIPS. tional Environmental Satellite References System Jones, A. S., and T. H. Vonder Haar, 2002: A dynamic parallel NPP NPOESS Preparatory Project data-computing environment for cross-sensor satellite data merger and scientific analysis. J. Atmos. Oceanic Technol., 19, OSDPD Office of Satellite Data Processing 1307-1317. and Distribution Kidder, S.Q., A. Jones, 2007: A blended satellite total precipi- SPC Storm Prediction Center table water product for operational forecasting. J. Atmos. Oceanic SSM/I Special Sensor Microwave Imager Technol., 24, 74-81. Randel, D. L., T.H. Vonder Haar, M.A. Ringerud, G.L. SSMIS Special Sensor Microwave Imager/ Stephens, T.J. Greenwald and C. L. Combs, 1996: A new global Sounder water vapor dataset. Bull. Amer. Meteor. Soc., 77, 1233-1246.

Fellowships in Atmospheric Science and Related Research The Cooperative Institute for Research in the Atmosphere at Federal and Colorado State laws, regulations, and executive orders Colorado State University (CIRA) offers a limited number of one-year regarding affirmative action requirements. In order to assist Colorado Associate Fellowships to research scientists including those on sabbatical State University in meeting its affirmative action responsibilities, ethnic leave or recent Ph.D. recipients. Those receiving the awards will pursue minorities, women and other protected class members are encouraged their own research programs, collaborate with existing programs, and to apply and to so identify themselves. The office of Equal Opportunity participate in Institute seminars and functions. Selection is based on is in Room 101, Student Services Building. Senior scientists and quali- the likelihood of an active exchange of ideas between the Fellows, the fied scientists from foreign countries are encouraged to apply and to National Oceanic and Atmospheric Administration, Colorado State combine the CIRA stipend with support they receive from other sources. University, and CIRA scientists. Salary is negotiable based on experience, Applications for positions which begin January 1 are accepted until qualifications, and funding support. The program is open to scientists the prior October 31 and should be sent via electronic means only to: of all countries. Submitted applications should include a curriculum Professor Thomas H. Vonder Haar, Director, CIRA, Colorado State Uni- vitae, publications list, brief outline of the intended research, a statement versity, humanresources@cira. colostate.edu. Research Fellowships are of estimated research support needs, and names and addresses of three available in the areas of: Air Quality, Cloud Physics, Mesoscale Studies professional references. and Forecasting, Satellite Applications, Climate Studies, Model Evalua- CIRA is jointly sponsored by Colorado State University and the tion, and Economic and Societal Aspects of Weather and Climate. For National Oceanic and Atmospheric Administration. Colorado State more information, visit www.cira.colostate.edu. University is an equal opportunity employer and complies with all

11 REGIONAL IMPACTS Regional Impacts of Oil and Gas Development in the Western United States Marco A. Rodriguez1, Michael G. Barna2, Tom Moore3, Kristi Gebhart2, Bret Schichtel2, Bill Malm2 1 Cooperative Institute for Research in the Atmosphere, Colorado State University, Fort Collins, CO 80523-1375 2 Air Resources Division, National Park Service, Fort Collins, CO 80523-1375 3 Western Regional Air Partnership, Western Governors’ Association, Fort Collins, CO 80523-1375 High ozone levels at the Ambient Air Quality Standards on air quality in rural regions.

earth’s surface, such as the (NAAQS) for ozone to 75 ppb For example, NOx emissions photochemical smog that (assessed as the 4th highest from an internal combustion frequently envelopes Los monitored ozone concentration engine at a gas well may react Angeles in the summer, have value over a running average with terpenes (a reactive VOC) typically been regarded as an eight hour period, over 3 con- emitted from pine forests and urban air quality problem. A tinuous years) from the previ- form ozone in an area where, disturbing trend in recent years, ous 80 ppb, effectively reducing previously, the right mix of however, has been the rise of the compliance level of the precursors was not available tropospheric ozone in remote ozone NAAQS by 9 ppb. Many for this reaction to take place. regions of the western U.S. medical experts assert, however, This is especially worrisome

Ozone (O3) is a strong oxidant that an ozone standard between since recent observations that can harm human health at 60 and 70 ppb is required to indicate that many remote Figure 1. Typical NOx emissions relatively low concentrations. protect human health. wilderness areas and national [moles NOx s-1] from oil and Recently, the U.S. Environmen- Ozone is formed through parks, such as Mesa Verde gas development in the western United States from the 2002 tal Protection Agency (EPA) a complex series of chemical National Park in southwest- WRAP emission inventory. tightened existing National reactions involving nitrogen ern Colorado, are confronted

oxides (NOx) and volatile with ozone concentrations organic compounds (VOC) that are trending towards the in the presence of sunlight. EPA’s acceptable limits. Very To combat rising ozone near Mesa Verde National Park levels, these precursors must are rapidly growing oil and be reduced. As oil and gas gas extraction operations in development in the western northwestern New Mexico. U.S. continues to accelerate, As this type of development however, there is significant continues throughout the West, potential that emissions from it is essential to understand its these sources will exacerbate potential negative impact on air the existing ozone problem. quality in some of our nation’s Although emissions from oil most cherished protected areas. and gas development may To investigate this issue, appear small as compared to CIRA and National Park Ser- other emission categories such vice (NPS) scientists are using as coal-fired power plants sophisticated meteorologi- and automobiles, they typical and air pollution models cally occur in remote regions to simulate air quality in the of the country, far removed western U.S., and determining from urban areas, and can which emission sources (e.g., have a disproportionate effect oil and gas development, power

12 plants, automobiles) contribute chemical processes that govern earlier version to the pollution burden in our the fate of pollutants. MM5 of the WRAP oil national parks and wilderness provides the wind fields that and gas emis- areas. This understanding is CAMx needs to determine the sion inventory, crucial to developing effective transport of chemical species, it is anticipated strategies that reduce regional as well as other meteorological that the general air pollution. Although this variables such as temperature trends presented article focuses on the impact of and mixing height. A detailed here give a gross ozone pollution, the concept of emission inventory focused indication of “one-atmosphere” computer on pollutants important for the impact of modeling is being employed by regional haze and visibility this source cate- CIRA and NPS researchers, and that covers the model domain, gory on regional collaboratively by the West- which includes the contigu- ozone forma- ern Regional Air Partnership ous U.S., southern Canada, tion. Future (WRAP) in their air quality and northern Mexico, specifies simulations will analyses and planning efforts1. the hourly flux of emissions incorporate the This approach is being used to from numerous area and point updated oil and investigate several issues related pollutant sources. The inven- gas emission to regional formation and tory consists of 22 emission inventory from transport of air pollutants such categories (e.g., automobiles, WRAP. as the primary and secondary powerplants, forest fires, oil CAMx was NAAQS for ozone and particu- and gas development) and used to simulate late matter, visibility protection, was developed for the WRAP. the entire year and mitigating health and eco- Figure 1 shows example NOx of 2002, which system effects due to excessive emissions associated with oil corresponds to nitrogen deposition and toxic and gas development in the the “base mod- air pollutants such as mercury. western U.S. Note the sig- eling year” being investigated by Figure 2a-b. Ozone time series Three major components nificant emissions that occur the WRAP, and the latest year for (a) simulated (black) and observed (red) ozone [ppb, 24 comprise the modeling system: throughout the Intermountain in which detailed emissions hour average] at Mesa Verde MM5 (Mesoscale Model 5), West, and in particular the Four were readily available. The first National Park during the a regional weather model, Corners region of northwestern step in this evaluation is the 2002 base case simulation, CAMx (Comprehensive Air New Mexico. comparison between predicted and (b) the change in ozone Quality Model with Exten- The oil and gas emission ozone concentrations with concentration [ppb, 24 hour average] at this site due solely to sions), a chemical transport inventory used in this study available observations. Once VOC and NOx emissions from model, and an inventory of was initially compiled for the the performance of this “base oil and gas development. pollutant emissions. CAMx WRAP’s regional haze simula- case” simulation is deemed simulates the emissions, disper- tions, with a focus on NOx and adequate, a second simula- sion, chemical reactions, and oxidized sulfur (SOx) emis- tion that evaluates the impact removal of pollutants in the sions, which are precursors of oil and gas emissions is troposphere by solving the to fine particulate nitrate and performed. Ground-level ozone pollutant continuity equa- sulfate, respectively. However, concentrations predicted by tion for each chemical species subsequent versions of this CAMx were compared to sur- on a system of nested three- inventory have been developed face measurements at 22 sites dimensional grids. Although and improved, and emissions within the Clean Air Status and computationally expensive, this of some species, such as VOC, Trends Network (CASTNET) type of simulation accounts have been substantially revised. monitoring network. The sites for the complex physical and Although this study uses the (continued on page 28)

13 HCEIRAADLIN commE uniqué

Correction to Fall 2007 Congratulations to NOAA’s workings of the global ocean 1995 Assessment. Graeme Communiqué 2007 Presidential Rank and atmosphere. In 2003, when Stephens (CIRA director- OAR was without an assis- select) was another important In June of 2007, the Award Winners tant administrator, he led and contributor. CloudSat SDPC Team received Each year the President organized a group of labora- a NASA Honor Award (Group recognizes a distinct group of tory directors to serve as acting 2007 Bronze Award Winners Achievement Award) for excep- career senior executives with assistant administrator (AA) Message From the Under Sec- tional contributions to the the President’s Rank Award for and deputy assistant adminis- retary of Commerce for Oceans CloudSat mission in the design, exceptional long-term accom- trator (DAA), and he served for and Atmosphere – development and implemen- plishments. High-performing two months as acting AA, and NOAA is pleased to tation of the CloudSat Data senior career employees are another two months as acting announce the winners of the Processing System. strong scientific leaders who DAA. Dr. MacDonald chaired 2007 Bronze Medal Awards. Michael Hiatt was also achieve results and consistently the Physical and Social Sciences Please join me in congratulat- in the list of team members demonstrate strength, integrity, Task Team (PSTT) group that ing the following individu- awarded. Michael developed industry, and a relentless com- looked at the entire organiza- als, groups and organizations the CloudSat hardware plat- mitment to excellence in public tional and geographic struc- for their dedicated service to form and DVD writing system. service. This year Dr. Alexander ture of NOAA’s research, and NOAA: Thank you for your dedication, E. MacDonald was one of two made recommendations that National Environmental and congratulations Michael! Presidential Rank Award win- were unanimously accepted by Satellite, Data and ners. Congratulations, Sandy! NOAA’s highest level of leader- Information Service Future Plans for CIRA Dr. Alexander E. “Sandy” ship. In August of 2006, he Ken Knapp MacDonald’s leadership and I am pleased to became OAR’s deputy assistant For analyzing terabytes of success in technology devel- announce that Graeme administrator for laboratories data from 18 different satel- opment and transition to Stephens has accepted our and cooperative institutes, lites from more than 20 years operations has helped make offer to serve as the next and the first director of Earth of observations in order to NOAA a leader in its opera- director of CIRA. He will System Research Laboratory create climate data records on tional systems. He is widely serve as CIRA director - (ESRL). hurricane trends and improve regarded as a visionary, who select beginning February 1, the understanding of climate is able to both predict where 2008. He will become CIRA Nobel Peace Prize Shared variability and change. science and technology are director on August 1, 2008. by CIRA Researchers Mark DeMaria, Antonio going, and to lead the way in Therefore, Tom Vonder Haar University Distinguished Irving, Nancy Merckle, John important programs. In the will remain CIRA Direc- Professor Tom Vonder Haar A. Knaff (Andrea Schumacher last five years, Dr. MacDonald tor through July 31, 2008 and Senior Research Scientist and Bernadette Connell were has made some extraordinary to allow for a transition as Emeritus Doug Fox worked also contributors). contributions to NOAA and we develop our proposal to with the Intergovernmental For the development and the Nation. He invented a new recompete for CIRA. Panel on Climate Change, operational implementation of display technology, Science On Please join me in which shared the Nobel Peace the Tropical Cyclone Forma- a Sphere, (SOS) that visually congratulating Graeme and Prize announced Oct. 12, 2007. tion product that quantitatively displays global data in a truly thanking Tom for many years Thousands of scientists from predicts storm formation prob- spectacular way. His new SOS of strong leadership. We truly across the globe were involved ability. display technology was awarded are fortunate to have two in the IPCC. Vonder Haar Conrad C. Lautenbacher, Jr. a patent to the government in such individuals within our made contributions to several Vice Admiral, U.S. Navy (Ret.) 2004. These popular edu- college. chapters, while Doug Fox was Under Secretary of Commerce cational systems are now in Sandra Woods, Dean the coordinating lead author for Oceans and Atmosphere multiple museums, resulting College of Engineering for the “Impacts of Climate and NOAA Administrator in the education of hundreds Colorado State University Change on Mountain Ecosys- of thousands of people on the tems Chapter” in the IPCC

14 HEADLINE

CIRA Scientist Among when you don’t know what the 2007 comes from Information participants during training. Authors of Book Celebrating weather is in lots of places,” Systems Branch Chief Carl She also assisted during train- Kidder said. “You’ve got to Bullock. ing and answered questions 50 Years of Earth see the big area to make the “I am nominating Leigh during lab exercises. These Observations from Space forecast. Satellites provide us Cheatwood-Harris for GSD’s contributions helped lead to a Stan Kidder has contrib- with that capability - looking at November 2007 Team Member successful training experience uted to a new book celebrating temperature and humidity all of the Month. Leigh has sup- for the participants, most of 50 years of Earth observa- over the Earth.” ported a number of important whom had little prior experi- tions from space. He is one Before satellites, severe evaluation and training activi- ence with the GFE.” of a dozen scientists from weather surprised people, ties in ISB over the past several around the country who have often resulting in thousands months. Two of particular note December 2007 GSD Team written chapters in the book of deaths. Kidder points to the were her work with the AWIPS Member of the Month: Earth Observations from Space: hurricane that hit Galveston, II development and GFE train- Sean Madine The First 50 Years of Scientific Texas, with no warning in 1900 ing for River Forecast Center The following nomina- Achievements. The book was and killed 8,000 people. “Now, forecasters. tion for GSD Team Member of sponsored by The National there are no surprise tropical For AWIPS II, Leigh was the Month for December 2007 Academies in honor of NASA’s storms anywhere on Earth,” he persistent in working with the comes from Aviation Branch 50-year anniversary in 2008. said. version 1.0 despite many prob- Chief Mike Kraus. Dr. Tom Vonder Haar Kidder’s research at CIRA lems typical of a first release “Sean Madine is GSD’s helped manage the book project focuses on using satellite data of software. She became the Team Member of the Month as a member of the Board on to study meteorological prob- de-facto expert on this system for December. We in the Atmospheric Sciences and Cli- lems such as the amount of helping others learn and use Aviation Branch would like mate. Vonder Haar also serves water vapor in the atmosphere, AWIPS II. She also developed to recognize Sean’s significant as the chairman of the interdis- which is useful for forecasting test metrics which will be used contributions to the Forecast ciplinary section of the National rain. He takes satellite informa- to evaluate the new AWIPS Verification Program. Sean is a Academy of Engineering. tion and builds products for the II system. This development key contributor to the success “The report sponsored by scientific community including included compiling and evalu- of the Verification Program. He the National Academies at the forecasters. ating usage logs, interacting provides programmatic coher- request of NASA is timely as A sample of Kidder’s and with NWS staff at the Boulder ence, project leadership and we plan the next segment of CIRA’s work: forecast office, and running vision, and has infused innova- our space missions to moni- • CIRA local-scale products: and documenting performance tive scientific concepts into tor Planet Earth,” Vonder Haar amsu.cira.colostate.edu/ tests. These metrics will form the verification and evaluation said. “Our past accomplish- GOES the basis of comparing the process. For instance, Sean is ments and lessons learned will • Water vapor imagery, includ- performance of the new AWIPS leading the design of the next assist with the design of future ing GPS data: amsu.cira. II system. generation RTVS, is developing Earth climate and science mis- colostate.edu/GPSTPW She has also played a key a significant cutting edge proj- sions.” • Oceanic water vapor imagery role for developing and testing ect with Boeing Corporation to Kidder talked about his around the world: amsu.cira. training material for initial understand the economic value chapter on the contributions colostate.edu/TPW GFE training conducted with of weather forecasts, and is satellites make to weather fore- River Forecast Center forecast- working to extend verification casting on February 17 at the November 2007 GSD Team ers from across the nation. She concepts and information for American Association for the Member of the Month: meticulously went through automated decision support.” Advancement of Science annual Leigh Cheatwood-Harris training material before “Thanks Sean for all you meeting in Boston. The following nomina- training began and suggested do and congratulations!” “People have been inter- tion for GSD Team Member changes to the material in order ested in weather for hundreds of the Month for November to make it more meaningful to of years, but you can’t forecast

15 CIRA communiqué

Announcing the 2007 GSD Please congratulate the topics and develop and deliver past few years, she has assumed Web Award Winners – following employees on teletraining to students – pri- increased technical leadership marily to NWS forecasters as roles and played a significant The “Webbies!” their recent promotions: well as other NOAA and DOD part in several collaborative Best New Site Christopher Anderson personnel. projects with ESRL/GSD and Jeff Smith Chris was promoted to Robert DeMaria NWS. research associate III in January “Data Locator” Site On July 1, Robert began Kathy Fryer 2008. He has been a researcher This new site provides a a half-time research associate In recognition of Kathy’s at CIRA in Boulder for several search engine for finding GRIB I assignment at CIRA in Fort evolving role within the years and has applied his atmo- and NetCDF meteorological Collins, a promotion from his RAMM Branch at CIRA in Fort spheric modeling expertise in a data here at GSD. Employing former assignment as a coor- Collins, she was promoted to a wide variety of endeavors that a web coverage service (WCS) dinator (non-student hourly). research coordinator position the ESRL/GSD/FAB (Forecast to return subsets of desired He is a familiar face here at in November 2007. The promo- Applications Branch) has been data in NetCDF format, the CIRA, as he has worked part tion recognizes the broad range involved in. Chris’ Hydrom- Data Locator web application time with the RAMM Branch of assignments for which she is eterological testbed (HMT) can either return this data as a for the past seven years – first responsible and her indepen- work has numerous compo- downloaded NetCDF file (.nc), as a high school PACE student dence in carrying out most of nents with the real time model or it can generate web pages and then while he completed those duties. Congratulations forecasts/ensembles, and he for viewing the data in a web an undergraduate degree in to Kathy! coordinates his efforts together browser or from within Google Computer Science at CSU. In with those of two other FAB Hiro Gosden Earth. A web service is also his new position, Robert will researchers in HMT-related Hiro was promoted to available for other applications work with database formatting ensemble post-processing. research associate III in August to invoke in order to find and and conversion, development Chris’ other area of climate- 2007. He is responsible for retrieve this data. of graphical display programs, related work takes the form of computer support to the Best Product/Internal Use algorithm verification pro- WRF downscaling with LAPS NOAA RAMM Branch at CIRA Patrick Hildreth grams, management of remote analyses for the ESRL Carbon in Fort Collins. This includes “FIDO” – Short for FICS Docs sensing databases, and code Cycle group, as well as interac- Windows and Linux system The GSD ITS Opera- optimization. In his off-time, tions with the NOAA Western administration, hardware tions staff have long relied Robert enjoys the challenge of Water Assessment. repairs, network support, and upon FICS to assist them in programming computer games. student hourly supervision. He Daniel Bikos monitoring the various systems Bernie Connell is Robert’s also serves as backup adminis- Dan works with the in support of the research supervisor. trator for the RAMMB AWIPS RAMM Branch at CIRA in projects here at GSD/ESRL. Joanne Edwards system. In addition, Hiro main- Fort Collins and was promoted “Fido,” short for FICS Docs, Joanne was promoted to tains various meteorological to a research associate III in is an add-on to FICS used to research associate IV in July workstations utilized by NOAA July 2007. Within the RAMM store and manage the detailed 2007. She has been a researcher and CIRA researchers to ingest Branch, Dan has worked with information pages associated at CIRA in Boulder since global satellite data and develop the Virtual Institute for Satellite with these systems. With Fido, February 1996 and a research research products. Hiro serves Integration Training (VISIT) operators are now able to edit, associate III since the AP career as manager for several projects, program since its inception add, and delete information ladder was implemented in including a current project in 1998 and has helped VISIT pages as well as view an archive 2002. She received the CIRA to streamline the process for become established as a pre- of previous versions. The Research Initiative Award in research product development, mier virtual training program primary users of the system 2000 as a key member of the evaluation and transition to for the NWS. He collaborates like it so well that they have Department of Commerce operations. with colleagues from CIRA, expressed a desire to move all Gold Medal-winning AWIPS CIMSS, and the NOAA/NWS of their documentation to this development team. Over the new system. Training Branch to research

16 Robert Hale Isidora Jankov tist III in January 2008. Over own web pages in an efficient Dr. Hale (located in New Dr. Jankov joined ESRL/ the past 2 years, Dr. Jiang manner. He coordinates every Mexico) was promoted to GSD’s CIRA staff in Boulder as has become an indispensable aspect of the development, research scientist I on Decem- a research scientist on Decem- member of a research collabo- maintenance, and administra- ber 1st, having served CIRA ber 1st. Isidora first joined ration involving the Chemical tion of the RAMMB website as postdoctoral fellow for GSD’s Forecast Applications Sciences Division at NOAA/ and research databases. Kevin three years. Dr. Hale’s research Branch at CIRA in Boulder in ESRL. Among her accomplish- works closely with NOAA web focuses upon changes in February of 2006 as a post- ments are her conversion development staff to ensure observed temperatures associ- doctoral fellow working on of the RAMS model to run that all NOAA security issues ated with changes in land cover ensemble modeling and rainfall on NOAA’s parallel process- and standards are addressed. at climate station locations forecasting efforts. Isidora ing supercomputer which In addition to his web master throughout the conterminous has her Bachelor of Science has permitted the group to responsibilities, Kevin provides USA. His supervisor is Tom in meteorology from Bel- simulate their very detailed a wide variety of Windows Vonder Haar. grade University in Serbia and microphysical models in 3-D hardware and software sup- Paul Hamer Montenegro, and her Master of and over significant simulation port to research staff. Addi- Paul was promoted to Science and Ph.D. in meteorol- times. She has also played a tional responsibilities include research associate IV in July ogy from Iowa State University very significant role in expan- system configuration, repair, 2007. Paul has been a researcher in Ames, Iowa. She remains sion of their coupled model to and new hardware and soft- at CIRA in Boulder since with the Forecast Applications include an aerosol model that ware research, evaluation, and 1998 and a research associate Branch – Development Section interacts with the dynamical, implementation. Kevin also III since the AP career ladder involved in research activities microphysical, and land-surface supervises two student hourlies was implemented in 2002. that span improvement of the modules. This allows the group assisting with web page devel- During this time, Paul has been diabatic analysis methodol- to see aerosol effects on clouds opment. responsible for the redevelop- ogy, evaluation of model error as part of a coupled system Karen Milberger ment of the Central Facility’s and error covariances, assess- with multiple feedbacks. She Karen was promoted data acquisition and processing ment and implementation has also been responsible for to research associate III in system. The implementation of 3-dimensional variational the huge task of simulating case July 2007. Karen has been a of the Object Data System he methods, ensemble design and studies from NOAA’s TEXAQS/ researcher at CIRA in Boulder totally designed and developed optimization for specific prob- GoMACCS (2006) field pro- since 2003. She received the has permitted GSD’s Informa- lems, ensemble post-processing gram in Houston. CIRA Research Initiative Award tion and Technology Services using advanced statistical Kevin Micke in 2004 as a key member of to more easily acquire, process, approaches, and configuration Kevin was promoted to the GLOBE Program. In 2006, store, and access a greater of probabilistic products for research associate II in Janu- Karen was handpicked by the amount of data. Paul also com- operational use. The culmina- ary 2008. He is responsible GLOBE Director to lead an pletely developed the long-term tion of her effort will be an for RAMMB web page devel- across-the board redesign of archiving of data using Facility experimental system aimed at opment and maintenance at their entire website. As the new Data Repository. Paul takes the supporting field efforts of the CIRA in Fort Collins and works Tech Lead for Web Support, initiative to explore new ideas Hydrometeorological and Haz- closely with research staff to she has been responsible for a and concepts to solve GSD’s ardous Weather Testbeds. transition real-time research major long-term effort to over- data acquisition and processing Hongli Jiang products to the web. The haul the software architecture challenges and to implement Hongli, a CIRA research RAMMB web page plays an to improve student-friendly novel approaches to maintain scientist II at CIRA in Boul- increasingly important role in visualization and to support the Lab on the forefront of IT der for over two years and a experimental product dissemi- easier navigation and interac- technological advances. research scientist with the CSU nation and evaluation. Kevin tive content. Department of Atmospheric has also developed template (continued on page 29) Science for over 10 years, was scripts that allow researchers promoted to research scien- to more easily develop their

17 HcIRA’EADLINS NEEW FACE

Vonder Haar, our department director, had the CIRA Has a Brand New Look idea last summer to renovate the CIRA building Marilyn Watson to give us a fresh look as we face the recompeti- CIRA has had a face lift! The final of seven tion for NOAA funds that is coming up this year. phases of renovation are now complete. Over Mary McInnis-Efaw, our CIRA manager and We have now completed 50 staff members have been jostled around as assistant director, asked Walt Naylor and me to a renovation project that the building has been worked on piece by piece. assist with the remodel along with two hourly has pulled everything Along with staff moves, all furniture has been workers, James Ubillos and Chris Musgrave. together into one unified moved; computers have been powered down, James and Chris were the backbone of all the moved, and powered up again. Phase after phase preparation and facilitating that went into pack- color scheme and given us it has gone on, and thus CIRA has gained a great ing up staff offices, settling folks into temporary a clean, fresh background new start. quarters and finally moving them back into their upon which to mount our The CIRA building is a unique building on freshly renewed office spaces. scientific canvas. the foothills campus of CSU. The first wing was The extensive remodel plan included new built in 1981, the second and third phases added paint inside and out as well as new carpet in 1987, phase four was completed in 1998, and throughout the building; new floor tile in the then came the main entry remodel of 1999. restrooms and break room and new ceiling tile in Because the building was constructed in several the older sections of the building. Also included phases, the carpet and paint were different in each were new half-light doors at upper and lower wing. Design and color choices are constantly exits, replacing dark solid doors to give the build- changing, making it difficult to match new to ing more light and open up the space a bit. old. We have now completed a renovation project We began the facelift mid-October of 2007. that has pulled everything together into one As the project progressed, more ideas for updates unified color scheme and given us a clean, fresh emerged. For example, over the years, computer background upon which to mount our scientific LAN lines were added in a constantly changing canvas. configuration as technology advanced and our The CIRA Renovation has been an interest- research staff was moved about to accommodate ing challenge for our department. Dr. Thomas project group changes. Because of the fluid needs

18 HEADLINE

of this ever-changing connection landscape, lines During this project it has been necessary for were originally run in the ceilings and dropped occupants of the CIRA building to be moved in down into whichever area needed at the time. phases to temporary locations either in our build- This proved to be an efficient if unsightly solu- ing or in adjacent buildings. The cooperation of tion. Now we have afforded ourselves the luxury the staff has been phenomenal! Complaints were of permanent LAN line jacks to allow connectiv- few and morale was high as everyone got into ity from most areas in each office. the spirit of the updated look. Kudos to CIRA In addition, broken seals were revealed on staff for tolerating amazing inconveniences and a great number of exterior windows which will continuing to produce their fantastic scientific have to be replaced. This will occur in the spring products in spite of the chaos of the remodel! of 2008 after the worst of the winter cold has given way to warmer weather. The renovation also gave us the opportunity to clean blinds and windows inside and out as well as clearing the dust from behind furniture that had not been dislodged for some time. We took this opportunity to clean the air ducts to rid the building of allergens lurking in the ceiling and wall cavities. Other changes include new handi- cap openers on our entry doors, new sinks and cabinets in the restrooms, a new chaise lounge in one restroom, new entry floor mats, and most conspicuous of all, a new counter top, all new conference tables and chairs, and a new wireless conference phone in our Director’s Conference Room. All of this has enhanced our building and our outlook.

Vonder Haar Celebrates 100th Student Since his arrival at Colorado State University in 1970, Dr. Tom Vonder Haar has enjoyed advising and mentoring more than 100 students to receive more than 100 degrees, including 76 M.S. and 27 Ph.D.s. Of course, some students obtain both the M.S. and Ph.D. in atmospheric science at CSU, so the question in the hallway has always been “Who will be Professor Vonder Haar’s 100th student to graduate?” The counting began easily with Tom’s first M.S. student (Capt. Gerry Sikula, 1971) and Ph.D. (Dr. Jim Ellis, 1975). Then, with co-advising of some and a few M.S. students taking the non- thesis option, the counting became a bit fuzzy. By combining the department records provided by Jen Weingardt, the Vonder Haar research group records compiled by Holli Knutson and Tom’s own count, it has been determined that his 100th student graduate ( ± 1 or 2) is among a group of four M.S. advisees that completed their degrees in Fall-Winter 2006. They are: Kate Maclay, Becca Mazur, Kevin Donofrio, and Dustin Rapp. Tom hosted a good sized group at his “100th student/graduate” lunch!

19 HEDEUCADLINATIONE AND OUTREACH Environmental Education and Public Outreach Julie Winchester

Acts of conservation without the requisite issues that will arise in the next 15-25 years. We desire and skills are futile. To create desire and have yet to make the link between environmental skill, and the ‘community motive,’ is the task of health and welfare and the impacts of cumulative education. – Aldo Leopold 1944 individual actions. According to a National Environmental The NEEFT, in collaboration with Roper Education and Training Foundation (NEETF) Research, has conducted surveys to better under- report, “Understanding Environmental Literacy stand what Americans know about the environ- in America and Making it a Reality,” 95 percent ment and how they view emerging environmental of American adults think environmental educa- problems. Figure 1 displays topics posed in a tion should be taught in schools and 90 percent 2000 survey that addressed subjects ranging from believe that adults in the workplace should also energy, water, and air pollution, to habitat loss receive some kind of environmental educa- and more. Figure 2 displays topics posed in a tion. Ten years of survey results show that while follow up study in 2001 that tested knowledge of American adults value the environment, their energy topics. In both charts, the numbers in the comprehension of complex environmental issues far right column represent the percentage of the is limited. About 80 percent of Americans are sampled population who answered the question influenced by incorrect or outdated environ- correctly. There were some surprising results. mental information and just 12 percent of those Two thirds of the American public says they surveyed can pass a basic quiz on awareness of know a fair amount about the environment but energy topics. At a time when we are confronted statistics tell a different story. Surveys show that Figure 1. with increasingly complex environmental choices, 80 percent of the public are heavily influenced by Percentage of sampled population who answered these our “EQ” (environmental intelligence quotient) incorrect information and it seems that we are basic environmental knowledge is extremely low and most of us are ill-prepared stuck in the environmental mindset of 30 years questions correctly. to understand and address the environmental ago. For example: Source: NEETF & Roper, 2001[1] • Only 27 percent of Americans know that 60 percent of our electricity is produced by burn- ing coal and other flammable materials. 40 percent of people believe that hydroelectric power is America’s top source of energy (only 10 percent of the total) • Add up hydropower, nuclear, and solar sources, and the majority of Americans think our electricity is generated in ways that have little or no impact on air quality • Just 22 percent of Americans know that runoff from farm fields, roads, parking lots, and lawns is the leading cause of water pollution in America today. 47 percent think the most common form is waste dumped by factories • 120 million Americans believe spray cans still have CFCs in them even though CFCs were banned in 1978

20 HEDEUCADLINATIONE AND OUTREACH

• 45 million Americans think the ocean is a source of fresh water • One in four Americans (23 percent) knows that paper products are the number one source of landfill material in America. 29 percent still think that disposable diapers are the greatest cause of over-stuffed landfills • Two-thirds of Americans fail to recognize that the transportation sector is the largest petro- leum user in the U.S. These facts and figures point to the need for comprehensive environmental education. While the American public continues to engage in basic activities (recycling, conserving electricity, and saving water) to benefit the environment, they are not embracing the fact that the lion’s share of our pollutants is generated by individuals and small entities. The environmental impact Figure 2. of small point sources is complex and is largely role of the scientist and researchers in education Percentage of sampled beyond the reach of environmental regulation and public outreach should not be ignored. population who answered programs, yet our environmental future depends Many research and funding agencies now these basic energy knowledge questions correctly. on personal knowledge and action. An informed require the integration of science and education Source: NEETF & Roper, 2001 [2] and knowledgeable public can play a larger role with mandates for greater involvement in public in evaluating proposed environmental laws and outreach. The role of the scientist in education determining new policies. Awareness of environ- and public outreach has evolved to more than just mental systems and an understanding of emerg- the K-12 school visit or the occasional public lec- ing environmental issues will spur the public to ture. One can advocate, be a resource, or join in a action in making long-lasting behavioral changes. partnership with educators and public outreach specialists. Scientists and Researchers – Working in tandem with education and Bridge to the Future public outreach specialists, scientists, and researchers brings: As we move into the 21st century, it is • strong content knowledge, essential that each of us has a sound knowledge • access to resources, of scientific concepts, processes, and technologies Figure 3. • critical and analytical thinking, in order to make informed decisions and take Roles for scientists and • scientific literacy, an active role in advancing our society. With a researchers. • technical support for outreach products. wealth of information available over the internet Source: Cherilynn Morrow, The following Venn diagram, conceptualized and through the media, our skills in information Space Science Institute, May by Cherilynn Morrow of the Space Science Insti- 2000 management, critical thinking, and problem- solving techniques will be sorely tested. The communication of sound research is increasingly important as we make choices about energy providers, support new sources of power generation, choose health care providers, vote to support stem cell research, or make behavioral changes that will lessen our impact on the environment. The vital

21 HEDEUCADLINATIONE AND OUTREACH

tations of scientific concepts behind air quality research, and interactive displays relating to air quality issues in national park visitor centers. In the public outreach arena, scientists reach out to the general public, policy makers, fellow scientists, and educators with presentations of their work. CIRA-NPS Web Presentation Nitrogen Emissions, Atmospheric Processes and Deposition Education and outreach programs at CIRA build on the expertise of scientists involved in the study of severe weather, weather forecasting, climate change, and air quality-related issues including visibility and acid deposition. The air quality research group depends heavily on local and national partnerships to carry out its goal to preserve, protect, enhance, and understand air Figure 4. tute, presents a framework for planning education quality and the relationship to sensitive resources Three circle Venn diagram in national park systems (Figure 5). offering a framework for and public outreach programs associated with planning education and scientific research programs (Figure 4). As one In an ongoing effort to develop outreach public outreach and programs moves from formal education, through informal materials that reflect current research activities associated with scientific education, and into public outreach, the capac- within the NPS air quality group, we are launch- research programs. ity to reach larger audiences greatly increases. ing a web-based presentation that seeks to convey Source: Cherilynn Morrow However, the potential to communicate in-depth an understanding of the sources and effects of Space Science Institute, May 2000 knowledge of the topic is greatly reduced. At all reactive nitrogen (both naturally occurring and levels there is ample opportunity to reach out to human caused) in the environment. (The term all age and educations levels with reactive nitrogen refers to all biologically active, shared knowledge that will per- photochemically reactive, and radiatively active haps inspire curiosity and further nitrogen compounds in the atmosphere and exploration of the topics. Many of biosphere of the earth). It focuses on the forma- the outreach products created by tion of reactive nitrogen in the atmosphere and the CIRA-NPS air quality research how it is transported and deposited in remote group at Colorado State University national parks and wilderness areas. Finally, it fit into the education and public explores a range of negative environmental effects outreach sections of this frame- caused when there is too little or too much reac- work. Under formal education, the tive nitrogen introduced into sensitive ecosys- “VIEWS” website provides real-time tems. The goal is to provide a resource for the access to monitoring data with soft- environmentally concerned community including ware tools, allowing users to create public officials, students, teachers, and the general summaries on demand. Undergrad- public. We focus on optimizing the link between uate and graduate students work education and research, taking care to present Figure 5 directly with researchers conducting field moni- facts and concepts, not opinions or advocacy. A Conceptual diagram of the toring and data analysis for special focus research The following highlights the web presentation on partners involved in CIRA- nitrogen deposition: NPS air quality research and studies. Informal education encompasses field the available public outreach training guides on CD-ROM, web-based presen- The most abundant gas in the atmosphere products. is nitrogen (N2). Today, human activities associ-

22 HEADLINE

ated with food production and fossil fuel com- While nitrogen compounds are bustion are increasing the amount of nitrogen important to the nature and diver- cycling between the atmosphere, soils, and water sity of plant life; excessive amounts resources. Reactive nitrogen is found both in its have deleterious effects on sensitive oxidized and reduced forms. Combustion sources, remote ecosystems including: such as power plants and automobiles, produce • an increase in global concentra- nitrogen oxide (NO) and nitrogen dioxide (NO2), tions of nitrous oxide (N20) a while fertilizing activities and feedlots produce potent greenhouse gas; ammonia (NH3), a reduced form of nitrogen. • increasing concentrations of nitric oxide (N) which drives the formation of photochemical smog; • acidification of soils and the loss of nutrients such as calcium and potassium necessary for soil fertil- ity; • acidification of lakes and streams including transport of nitrogen into sensitive coastal waters and estuaries; • loss of biological diversity in plants and microbes which impact In its molecular form, nitrogen is unus- the aquatic and animal species able by most living organisms. It must be trans- dependant on them to survive; formed into reactive nitrogen before plants and • lower numbers of nitrogen-fixing animals can make use of it. The application of organisms (algae and bacteria) reactive nitrogen, in the form of synthetic fertil- that convert gaseous nitrogen to izers, allows for increased food production but plant-usable forms; increases the release of nitrogen into the environ- • reduced visibility. ment through agricultural runoff and sewage. Public education about reactive nitrogen as an environmental problem is an essential first step towards making informed decisions to mitigate it. This presentation com- municates in nontechnical, understandable language: 1) the sources and storage mechanisms of reactive nitrogen; 2) the chemistry and flow paths it follows in the environment; and 3) the consequences for people and ecosystems as reactive nitrogen deposits in the environment. [1] Kevin Coyle, Report: “Environmental Literacy in Amer- ica,” 2005 The National Environmental Education & Training foundation, Washington, D.C. [2] Dr. Cherilynn A. Morrow, “A Framework for Planning Education and Public Outreach Activities Associated with Scientific Research Programs,” Space Science Institute, Boulder, Colorado www.spacescience.org/ click on “Papers on EPO.”

23 HDPEADLINC HARDWARE E Cloudsat DPC Hardware Michael Hiatt When CIRA started discussing a plan to assembles these systems onsite. System design build and operate the CloudSat Data Process- and spare parts are kept in CIRA’s control, greatly ing Center (DPC) with NASA/JPL, the Institute simplifying configuration changes and mainte- was already well established at data collection, nance. Furthermore, CIRA can perform in-place processing, distribution, and archive techniques upgrades should the need arise. These systems and technologies. CIRA has been operating a can accommodate two SATA hard drives, which satellite earth station since 1978. During this time, are typically configured in a redundant array CIRA has designed many unique hardware and of inexpensive disks (RAID) configuration for software solutions using mainstream technologies redundancy. Although the drives can store data, to improve and streamline its earth station. These they are configured primarily for the Operating improvements have significantly reduced costs System (OS). These systems perform all data col- and manpower needed to operate an earth station lection and processing. thus allowing CIRA to be very competitive when Meeting Internal and External Distribution bidding for data processing projects. Some of Requirements. Data storage and distribution is CIRA’s earth station design topologies were used handled by hardware dedicated systems called in the Cloudsat DPC to take advantage of this Network Attached Storage (NAS). Since a NAS efficient low-cost architecture. These mainstream device has the single purpose of data storage, it is technologies and their impact on the Cloudsat well optimized, using high density hot swappable DPC are discussed here. RAID bays, proper drive cooling, a high speed Limiting Physical Building Space Require- Ethernet port, embedded OS, and a web interface ments. The first consideration of any DPC is the with email notification for issues. NAS devices physical building space requirements. A dedicated are simpler to build and maintain since a general computer room is a primary requirement and purpose PC and OS are not required. Today, using has significant impact on the budget. The size, 1TB drives, CIRA’s NAS provides 3.6TB of RAID number, power requirements, and heat dissipa- storage for about $2,000. Heat dissipation, power tion of the computers dictate how the room will requirements, and footprint are similar to the be designed. CIRA has successfully used Small SFF PCs above. Performance is roughly equal to a Form Factor (SFF) Intel quad-core PCs with heat dedicated PC performing the same task. CIRA’s unique computer pipe technology in areas where space and cooling Robust Data Archive. Data archiving is the solution allows rooms to were limited and noise could be a factor. Com- last step in the process. In many DPCs, tape drives be used with less power puter racks using 19" rack PCs also offer high are the primary archive media. In 2001, CIRA and cooling, which keeps density solutions but double heat, power, and developed a DVD archive system that addressed construction costs low cooling requirements. CIRA’s unique computer the many shortfalls of tapes. These include lower solution allows rooms to be used with less power media cost, random access, reduced storage room and cooling, which keeps construction costs low. requirements, less expensive hardware to read Low-Cost Targeted Hardware Configura- and write, no special software to access, ability tion. The SFF PCs are purchased as bare bone sys- to verify without media degradation, and easy tems that only contain the chassis, power supply, duplication. An article in the Fall 2003 CIRA and main board. These are mainstream systems Magazine discusses this further. Since the system sold in mass quantity at competitive prices. CIRA was deployed, CIRA’s archive has performed at installs whichever Intel processor currently meets a 99.9 percent storage and retrieval rate. Due to the best performance vs. cost. Memory is installed the success of the DVD archive system, CIRA is using the amount needed for the task. CIRA able to archive more data than ever before. Today,

24 HDPEADLINC HARDWARE E

about 100GB of data is written and verified each easy to maintain, inexpensive systems which are day as a part-time task for one person. Room now well proven. CIRA continues to research and storage requirements have actually reduced as old develop new technologies that fit this model as tapes have been converted to DVD. they become available. CIRA is well positioned for In summary, use of mainstream technologies future DPC missions. customized for CIRA’s tasks has provided fast,

International Polar Year Expedition to the South Pole, Antarctica Dr. Glen E. Liston and Dr. Jan-Gunnar Winther Diary Excerpt (Glen), 1 December 2007 As I lie in my small tent at -48 °C (-54 °F) near the center of Antarctica, I try to mentally force my near-frozen toes back to life. The temperatures I am experiencing now are just about equal to the mean annual Antarctic air temperature in this area; something that typically changes only slowly over the decades and centuries. I also note that one month ago I was flying, uncontrolled, 15 meters (50 feet) through the air in response to a 52 m/s (132 mph) wind gust during a storm near the Antarctic coast. I am part of a scientific research expedition traversing from the Antarctic coast to the South Pole, and this is exactly what we are here to study and understand: weather and climate variability and change on time scales of 1000 seconds (like my wind-borne flight) to 1000 years (like my cold, cold night). Editor’s Note: Glen received a College of Engineering Out- standing Administrative Professional Staff Award on April 10, 2008. It was presented at the College of Engineering awards cer- “The story of this CIRA-sponsored International Polar Year (IPY) expedition will be featured in the Fall 2008 issue of the CIRA emony held at the Edna Griffin Concert Hall, Center for the Fine Magazine. In the article Dr. Liston will describe what it is like to Arts, Colorado State University. Remarks were made by Steven work in wind-chill factors as low as -69 °C (-92 °F), and to go 89 Miller, deputy director of CIRA, including excerpts from Glen’s days without a shower!” journal entries. Glen was unable to be there due to his travels.

25 NEW global model ESRL’s New Global Model and Its Implementation Jacques Middlecoff and Ning Wang

A new global weather Grid level Number of grid point Grid resolution (km) prediction model has been 6 40962 120.5854 developed at the NOAA Earth 7 163842 60.1623 System Research Laboratory (ESRL) in Boulder called the 8 655362 30.0806 Flow-following Finite-Volume 9 2621442 15.0408 Icosahedral Model (FIM) [1]. Table 1. Grid statistics for recursion levels 6 through 9. FIM is a hydrostatic model with a flow-following vertical For each grid point on the be many variations to this basic coordinate whose surfaces may sphere, a Voronoi cell is con- construction. It is possible to deform freely according to air structed around it. The Voronoi create different variations of flow. Aloft, the flow-following cell of a grid point is a point set the grid to meet the needs of coordinate is isentropic, while whose elements are closer to different numerical discretiza- near the surface, the coordinate the grid point than any other tion schemes or numerical surfaces are terrain-following grid points. Except for those computations of the dynamic [2]. The horizontal grid con- cells around the 12 starting grid operators. sists of hexagons that cover the points which have five edges, Grid Structure and world like the hexagons on a all other cells have six edges to Parallelization Figure 1. Icosahedral hexagonal soccer ball (see Figure 1). form a hexagon. The numerical Since FIM has an ico- grid and its Voronoi cell For more information on computations are carried out sahedral horizontal grid, a boundary. FIM, see http://fim.noaa.gov/. based on these cells (Figure 1). conventional rectangular grid Icosahedral hexagonal grid In general, it is a rather structure was not used. Instead, The horizontal grid mesh uniform grid mesh. The ratio the grid points are laid out in is created from a set of 12 of the greatest grid point a one-dimensional vector that evenly spaced grid points on distance to the smallest one is traverses the globe and adjacent the sphere. Each of the 12 grid less than 1.19512. The ratio of points are referenced by indi- points is connected to five the greatest Voronoi edge to the rect addressing (see Figure 2). neighbors to create 30 great smallest one is about 1.9. The Indirect addressing has impli- circle edges and 20 spherical relatively larger triangles appear cations for the parallelization equilateral triangles. The result- around the center of these 20 of the code. To distribute the ing polyhedron is called an starting triangles and smaller vector onto N processors, the Figure 2. Example of hexagonal Icosahedron. ones close to the corners of vector is simply divided into N nearest neighbors with not near these triangles. Table 1 lists indexes. From each of these equi- sections, one section per pro- lateral triangles, the following some statistics for the icosahe- cessor. Thus, adjacent points operations were performed dral grid for recursion levels 6 may be on another vector sec- recursively: find the middle through 9. tion or another processor and points of each great circle of A rather interesting aspect this affects how data is opti- the spherical triangle and con- of this construction is its sim- mally communicated between nect them with a great circle to plicity and regularity. It is easy processors. FIM is parallelized make 4 sub triangles [3]. The to analyze the geometric prop- using the Scalable Modeling resolution of the grid mesh is erties of the generated grid. System (SMS) developed at determined by the recursive On the other hand, depending ESRL’s Global Systems Division level specified (Table 1). on our aim or the metric we (GSD). SMS is a directive-based wish to optimize, there could system for parallelizing weather

26 and climate codes for distrib- communication optimization. Jacques Middlecoff, with Sandy uted memory machines. For Fortunately, an advantage of MacDonald and Steve Koch more information on SMS, see indirect addressing is that the advising, in the development CIRA Magazine Vol. 17, Spring code is independent of the data of a non-hydrostatic version of 2002 and http://www-ad.fsl. layout so the data layout can be FIM called the Non-hydrostatic noaa.gov/ac/sms.html. varied to see which data layout Icosahedral Model (NIM). NIM To accommodate the is optimal. We have developed will use a horizontal grid simi- indirect addressing of FIM, tools for automatically varying lar to FIM’s horizontal grid, but SMS was generalized to include the data layout and are just now will use the NOAA Geophysical general, indirect addressed, beginning experiments to see Fluid Dynamics Laboratory unstructured grids. We con- which layouts are optimal. The (GFDL) AM2 model’s vertical sider 2500 points per proces- first two layouts that will be Lagrangian coordinate system sor to be the “sweet spot” for tested are the Hilbert curve and for the vertical grid. NIM will FIM and, for 30 km resolution, a layout that mimics a Carte- also use a modified version of 240 processors results in 2731 sian coordinate layout. AM2 physics with the aim of points per process. Very good Some Results doing climate prediction as scaling (1.96X) is obtained In late February 2008, FIM well as weather prediction. Our from 120 processors to 240 began producing daily global 2-year goal is to run NIM at 3.5 processors. forecasts at 30-km resolution km resolution requiring about Grid Layout Optimization running on 240 processors. 31,000 processors. At that The path the vector actu- Thanks to extensive collabora- resolution, we hope to be able ally takes as it traces out all the tion throughout GSD, we were to perform a 2-month forecast points on the globe was not able to treat Mary Glackin, and predict the Madden-Julian specified above because the NOAA’s Deputy Undersecretary Oscillation (MJO). path is optional. Currently, a for Oceans and Atmosphere, Acknowledgements Hilbert curve is used to trace to a spectacular global weather The other members of the FIM out the path because, theo- forecast produced by FIM development team were Jin Lee (NOAA/ ESRL/GSD), Rainer Bleck (CIRES), Jian- retically, a Hilbert curve traces and displayed on Science on a Wen Bao (NOAA/ESRL/PSD), and Stan Figure 3. Precipitable water the points in such a way as to Sphere (SOS) during her visit Benjamin (NOAA/ESRL/GSD), with (cm) overlaid with geopotential maximize locality. Locality is to GSD in early March. Figures Sandy MacDonald (NOAA/ESRL) and height at 500mb. good for cache utilization, but 3 and 4 are two examples John Brown (NOAA/ESRL/GSD) in an advisory role. caches are very complicated. of what was shown to Ms. [1] Lee, J.L., R. Bleck, A.E. MacDon- Depending on the machine, Glackin. ald, J.W. Bao, S. Benjamin, J. Middlecoff, there can be up to three levels For more information on N. Wang, and J. Brown, 2007: FIM: A of cache, each level having SOS, see CIRA Magazine Vol. vertically flow-following, finite-volume icosahedral model. Preprints, 22nd Con- arbitrary size, bandwidth, 23, Spring 2005 and http://sos. ference on Wea. Analysis and Forecasting / latency, and replacement noaa.gov/. 18th Conference on Num. Wea. Pred., strategy. Additionally, the cache Future Plans 24-29 June 2007, Park City, UT, Amer. performance depends on the Meteor. Soc. FIM has now moved into a [2] Benjamin, S.G., G.A. Grell, J.M. number and size of the arrays testing phase under the leader- Brown, T.G. Smirnova, and R. Bleck, being used and on the data ship of Stan Benjamin. After 2004: Mesoscale weather prediction reference patterns. It’s clear that successfully completing testing, with the RUC hybrid isentropic-terrain- Figure 4. Wind speed (meter/ figuring out the best curve for FIM will become part of the following coordinate model. Mon. Wea. sec.) at 500mb and geopotential cache optimization is a daunt- Rev., 132, 473-494. height at 500mb. NCEP production ensemble. [3] Baumgardner, J.R. and P.O. Fred- ing task, further complicated by Jin Lee will lead Jian- erickson, 1985: Icosahedral discretization the fact that changing the curve Wen Bao, Ning Wang, and of the two-sphere. SIAM J. Numer. Anal., also affects inter-processor Vol. 22, No. 6, pp. 1107-1115.

27 HREADLINGIONALE IMPACTS continued from page 13

focus of this study) CAMx is This estimate was calculated by able to consistently predict the evaluating two CAMx simula- general annual trends for ozone tions: the base case simula- concentrations, with an annual tion, in which all emission normalized error and normal- categories are accounted, and ized bias of 22.7% and -1.6%, a “no oil and gas” simulation, respectively, falling well within which is similar to the base the EPA’s guidelines for accept- case, except that oil and gas able model performance. emissions are removed. The For this study, the larg- difference between these two est impacts from oil and gas predictions represents the emissions on regional ozone contribution of oil and gas were seen near Mesa Verde emissions on regional ozone. National Park. This is not Notable in Figure 2b is the surprising, given its proximity fact that oil and gas emis- to the extensive development sions can actually decrease that is occurring in northwest- ozone concentrations at Mesa ern New Mexico as shown in Verde National Park through Figure 1. Figure 2a shows the the process of “NOx scaveng- predicted (black) and observed ing”, where available ozone is (red) ozone concentrations consumed by reacting with for Mesa Verde National Park, nitric oxide (NO). This effect and Figure 2b illustrates the is most prevalent in the winter, change in ozone concentration when ozone concentrations at this site attributed to oil and are lower (at approximately 40 gas emissions. As expected in ppb), which is near the global Figure 2a, the general trend of background concentration. both predicted and observed However, in the summer, the ozone is low concentrations situation is reversed, and warm, during the colder winter stagnant conditions yield an months, when limited pho- increase in ozone from oil and tochemistry will occur, and gas emissions. Although these higher concentrations during impacts appear relatively small the warmer late spring and (e.g., an increase of a few ppb summer months, when meteo- in the summer), it should be rological conditions are more remembered that this period favorable to ozone production. corresponds with seasonally- In addition, it is anticipated high ozone concentrations, and that enhanced biogenic VOC that the results presented here chosen fall within the western emissions occurring during the represent a 24 hour average; Figure 3a-b. Peak predicted spring and summer will further on an hourly basis, the impacts annual ozone maxima [ppb, 1 region of the United States. hour average] in the western An evaluation of CAMx’s skill influence ozone formation in at Mesa Verde National Park U.S. from (a) the 2002 base in predicting ozone is done in this region. Figure 2b shows the can be much more significant, case simulation and (b) the accordance with the EPA’s sug- resulting change in predicted approaching 10 ppb. enhancement from VOC and ozone concentrations that are A regional perspective of NOx emissions from oil and gas gested performance guidelines development. for ozone modeling. In general, attributed solely to emissions ozone formation is illustrated within the western U.S. (the from oil and gas development. in Figure 3. Figure 3a shows

28 HEADLINE

the peak estimated ozone days with light winds in the ozone concentration is due New Mexico. Class I areas in concentration at each model summer, when the meteorol- solely to emissions from oil and this region that are likely to be grid cell that occurred during ogy is most favorable for ozone gas development. Although the impacted by increased ozone the 2002 base case simulation. production. These periods peak ozone maxima through- include Mesa Verde National As expected, there are high also typically correspond to out the West are typically Park and Weminuche Wilder- concentrations, exceeding 120 peak VOC emissions from quite small, there is a strong ness Area in Colorado, and San ppb, downwind of major urban biogenic and anthropogenic signature of a 2-3 ppb of ozone Pedro Parks Wilderness Area, areas such as Los Angeles, sources. The role of NOx and throughout New Mexico, Bandelier Wilderness Area, San Francisco, Salt Lake City, VOC emissions from oil and Colorado, and Wyoming, with Pecos Wilderness Area, and and Denver. There are also gas development on ozone in a pattern that approximates the Wheeler Peak Wilderness Area large ozone peaks in the more the western U.S. is shown in emissions shown in Figure 1. in New Mexico. remote regions of Nevada, Figure 3b. Similar to Figure In addition, significant ozone This study, although not Wyoming, Utah, Arizona, New 3a, the peak simulated ozone concentrations, exceeding 10 exhaustive, does indicate a clear Mexico, and Colorado. These for each grid cell during 2002 ppb, are evident in southwest- potential for oil and gas devel- maxima occur during hot is shown, but in this case, the ern Colorado and northwestern opment to negatively impact regional ozone concentrations in the western U.S., including several treasured national parks and wilderness areas in CIRA communiqué continued from page 17 the Four Corners region. It is likely that accelerated energy Yoo-Jeong Noh 2005, was promoted to research of verification metadata, a development in this part of the Dr. Noh was promoted in associate III in January 2008. component critical to the suc- country will worsen the exist- February 2008 to research aci- Primarily working on verifica- cess of the project. ing problem. The formation of ozone pollution examined here entist II. Having served CIRA tion of meteorological products Tong Zhu represents a complex phenom- in Fort Collins for two years as for use in aviation, she made Dr. Zhu was promoted to enon involving non-linear a postdoctoral fellow, she has significant contributions to the research scientist II on Septem- physicochemical processes, achieved significant research Real-Time Verification System ber 1st. Having served CIRA uncertain emission invento- results relevant to snowfall (RTVS), which currently serves for three and a half years as a ries, and fine-scale transport retrievals over land utilizing the needs of the NWS, the FAA, postdoctoral fellow located in in mountainous terrain. These high-frequency satellite micro- and the research community. Camp Springs, Maryland, he simulations will be refined with wave measurements. Recently, Missy has also worked on the has helped to develop a unique the updated emission invento- Dr. Noh’s study of remote sens- team tasked with designing and hurricane model initialization ries available from the WRAP. ing of snowfall has expanded demonstrating a verification scheme that allows impor- Although a daunting technical to incorporate microwave and service for use in the nation’s tant assimilation of satellite problem, regional air quality radar radiative model system future air traffic manage- microwave data into mesoscale modeling remains the only developments and soon she will ment system. The verification models. Currently, Dr. Zhu par- feasible option for developing interface with the CIRA/CSU software, called the Network- ticipates actively in a GOES-R emission control strategies that clouds/radiation/precipitation Enabled Verification Service AWG project and JCSDS OSSE have the potential to reduce research group and collabo- (NEVS), will allow integration study. rate with CIRA/CSU’s Center of verification data with other ozone concentrations and pro- for Geosciences/ Atmospheric information necessary to sup- tect air quality. Research, Army Research Labo- port air traffic flow algorithms. Harold Gibson [1] Western Regional Air Partner- ratory, and CloudSat scientists. Missy’s role has involved design Harold Gibson, ship 2008-12 Strategic Plan; March Research Scientist Emeritus, 2008; www.wrapair.org/WRAP/ Melissa Petty and proof-of-concept imple- documents/WRAP_2008-12_Strate- passed away January 5, 2008. Missy Petty, a researcher mentation for the management gic_Plan3_08final.pdf for CIRA in Boulder since

29 CIRA Mission The mission of the Institute is to conduct research in the atmospheric sciences of mutual benefit to NOAA, the University, the state, and the nation. The Institute strives to provide a center for cooperation in specified research program areas by scientists, staff, and students and to enhance the training of atmospheric scientists. Special effort is directed toward the transition of research results into practi- cal applications in the weather and climate areas. In addition, multidisciplinary research programs are emphasized, and all university and NOAA organizational elements are invited to participate in CIRA’s atmospheric research programs. The Institute’s research is concentrated in several theme areas that include global and regional climate, local and mesoscale weather forecasting and evaluation, applied cloud physics, applications of satellite observations, air quality and visibility, and societal and economic impacts, along with cross-cutting research areas of numerical modeling and education, training, and outreach. In addition to CIRA’s relationship with NOAA, the National Park Service also has an ongoing cooperation in air quality and visibility research that involves scientists from numerous disciplines, and the Center for Geosciences/Atmospheric Research based at CIRA is a long-term program sponsored by the Department of Defense.

Cooperative Institute for Research in the Atmosphere College of Engineering – Foothills Campus, Colorado State University 1375 Campus Delivery, Fort Collins, CO 80523-1375 (970) 491-8448 www.cira.colostate.edu

If you know of someone who would also like to receive the CIRA Magazine, or if there are corrections to your address, please notify us.