Trip Report - Joint IUM/NCAR Workshop on Urban Meteorology: Observing and Modeling

1.- Introduction

     The purpose of our trip to Beijing, People's Republic of China, was to participate in the Joint Institute of Urban Meteorology (IUM; a Chinese agency)/National Center for Atmospheric Research (NCAR) Workshop on Urban Meteorology: Observing and Modeling; and to visit the Chinese Academy of Meteorological Sciences (CAMS). Both IUM and CAMS fall under the umbrella of the Chinese Meteorological Administration (CMA). The report will detail the presentations given at the urban meteorology workshop, as well as briefly describe the visit to CAMS, and provide a description of some of the cultural experiences I had during my stay.

2.- Workshop

     Over the course of the Joint IUM/NCAR Workshop, there were a total of 13 presentations, not including opening/closing addresses, discussions, and summary presentations. The workshop was divided into three sessions: i) Observing Urban Boundary Layer (UBL), ii) Urban Landuse Modeling, and iii) Urban Air Quality and Pollution. After the presentations, the co-chairs of each session gave summary presentations on the highlights of their session, and finally, a discussion on possible collaborative efforts between Chinese and US meteorologists took place.

2.1 Opening Statements

     Mr. Pue Xie, Director of the Beijing Meteorological Bureau, officially welcomed all attendees of the conference, and turned the conference over to the chairs, Ms. Yingchun Wang, Director, IUM, and Dr. Fei Chen, Scientist, NCAR. Ms. Wang and Dr. Chen both detailed the need for significant involvement by U.S. and Chinese scientists to study urban meteorological issues. Factors such as homeland security and the necessity of understanding weather and environmental fields for emergency response lead the way in the U.S., while factors such as air quality are prime motivators for the Chinese. Dr. Chen also went over a few of the workshop objectives during his presentation, including: sharing research and development progress, identifying research and development priorities, and the exploration of U.S.-Chinese collaborations. Ms. Wang gave an overview of the organization of the IUM, including some of the IUM's recent accomplishments (e.g. implementation of a mesoscale NWP system in Beijing, and air quality forecasting system, etc.).

2.2 Session I: Observing UBL

     Mr. Ju Li of IUM and Dr. Jielun Sun of NCAR co-chaired the first session of the workshop. Dr. Sun gave the first presentation of the session, entitled "Current Issues in UBL Meteorology." In her presentation, Dr. Sun talked about many of the perturbations caused by an urban environment, including an alteration of atmospheric energy balance, atmospheric momentum balance, visibility and actinic flux, and precipitation and other storm processes. Additionally, she detailed several complications that arise in attempting to model an urban environment, including strong shear at the top of the urban canopy (UC), wake diffusion behind individual roughness elements, form drag, and reduced thermal stability, among others. In summary, Dr. Sun noted that: Monin-Obukhov similarity works only in the surface layer of an UBL; within the roughness sublayer, turbulence is dominated by buildings and surface structures, and is essentially site-dependent; and the nocturnal UBL can be dominated by heat release, leading to a neutral or slightly unstable BL.

     The next presentation of the session was given by Mr. Li, and was entitled "Introduction of Beijing UBL Experiment 2004 (BUBBLEX2004)." The motivation for BUBBLEX2004 was the necessity of increasing urban meteorological observations, with the ultimate goal of meeting the requirements of meteorological services for the 2008 Olympics, which will be held in Beijing. It was felt that research on the UBL and the urban heat island (UHI) would help to advance this goal. During BUBBLEX, observations of surface temperature distribution, wind field distribution, roughness, diurnal height of the UBL, and profiles of temperature, moisture, and wind in the UBL were made during the summer of 2004. For the experiment, over 40 automated weather stations provided by the Beijing Meteorological Bureau (BMB), 20 additional weather stations, three tethered balloon systems, three profiles, one lidar, one radiosonde station, one 325-m tower (with measurements at 15 different levels), and one flux station (eddy covariance system) were deployed. With the massive amount of data that were collected during the experimental period, efforts are underway to synthesize the information.

     Following Mr. Li's presentation, Dr. Don Lenschow of NCAR gave a presentation entitled "Modeling and Observing the UBL." Dr. Lenschow's presentation began by characterizing surface alterations in urban areas, and then drawing parallels between the UBL and a forest BL, including that both are characterized by a shear layer at the top of the canopy, weak diffusion, form drag, and organized coherent motions within the canopy that likely result from the mixing layer at the top of the canopy. This conclusion can aid researchers attempting to model or observe the UBL, because modeling and observing the forest BL is a field in which meteorologists have a little more experience. He also described some techniques for observing the UBL, including the use of laboratory-based tank and wind tunnel studies (these are generally idealized studies), remote sensing instruments, such as radars and lidars (these suffer sensitivity and resolution limitations), aircraft (these cannot do long-term observations), and arrays of instruments (which are limited by both the number of sensors and deployment difficulties). In conclusion, Dr. Lenschow outlined some of the unique problems associated with the UBL, including inherent heterogeneity and complexity, the fact that multiple processes are relevant (e.g. heat island, pollution, precipitation enhancement, etc.), and the fact that special observational problems exist, including observations at peak-traffic times and deployment limitations. He suggested that these problems might be overcome by a combination of observations, modeling (such as using large-eddy simulations), and laboratory simulations.

     Dr. Mingyu Zhou from the State Oceanic Administration was next to give a presentation, entitled "Urban Albedo Derived from Satellite Remote Sensing in Beijing Area." In this study, the Moderate Resolution Imaging Spectroradiometer (MODIS) satellite was used to derive albedo from the Beijing area. It was found that the lowest values of albedo appear in the center of the city, increasing with distance from the center. It was also determined that, in Beijing, the albedo is higher in the summer than it is in the fall, and that the variation of albedo is greater outside of the city than inside. It was concluded that such studies can be used to develop better landuse information as an input for the urban models.

     The final presenter of the first session was Professor Weimei Jiang from Nanjing University (NJU). The title of his presentation was "Research on UBL in NJU." The presentation provided an overview of studies undertaken or planned to observe and model the UBL, numerical simulation research in an urban meteorological environment, and application research in the urban atmospheric environment. While this presentation detailed the efforts of various projects, the crux of the presentation was that, with the results obtained from all of the studies undertaken, it should be possible to: i) build and improve observational and experimental setup in the urban inhomogeneous surface conditions, ii) utilize remote sensing techniques for UBL research; especially for investigating the interactions between dense groups of cities and the PBL under inhomogeneous surface conditions, iii) improve and develop UBL multi-scale modeling systems and iv) introduce the structures and characteristics of the UBL into regional forecasting models and climate models. Finally, Dr. Jiang highlighted some of the outstanding research issues for the UBL: i) how do urban areas affect weather, ii) how is mixing between street air and BL air mediated, and iii) how is weather 'felt' in urban areas.

2.3 Session II: Urban Landuse Modeling

     Drs. Chaolin Zhang of IUM and Dev Niyogi of North Carolina State University co-chaired this session, with Dr. Fei Chen of NCAR giving the first presentation, entitled "Recent Progress in Urban Land-Use Modeling for Community MM5 and WRF Models." Dr. Chen talked about the development, implementation, and testing of the community Noah LSM with an urban canopy model. He noted that the goals of urban landuse modeling for high-resolution NWP models are to provide more accurate weather forecasts, especially near surface and PBL structures, for urban regions, and to provide meteorological fields for air quality and dispersion models. He highlighted that many challenges exist in the quest to improve urban landuse modeling. These include: the specifications of urban landuse and urban parameters (such as fraction of vegetated and impervious surfaces, emissivity, albedo, and the geometry of buildings), the degree of complexity that should be taken in urban modeling (a simple approach versus a complex urban canopy model), and initializing the state variables of urban models (including such things as road, wall, and roof temperature). Work is currently being undertaken at NCAR to attempt to resolve some of these very issues. Summarizing some of this current work, Dr. Chen discussed upcoming capabilities of urban-canopy modeling, including providing detailed distribution of the urban heat islands, enhanced forecast of mesoscale wind and thermal structures over and around the urban areas, and providing more accurate input for air quality and dispersion models. Dr. Chen also noted that the initialization of urban models is still a major challenge, as there is high sensitivity to soil moisture fields, and that high-resolution land data assimilation needs to incorporate the urban canopy model (UCM) component. To this end, the UCM is being coupled to the WRF v2.0, and will be tested for New York City and Washington, D. C.

     The next presentation was by Professor Dui Wu of the Institute of Tropical and Marine Meteorology (ITMM) and Dr. Xuejian Deng, also of ITTM. The presentation was entitled "Scientific Experiment on Land-Atmosphere Interactions and Atmospheric Brownish Haze for the Cities in the Pearl River Delta". They presented details regarding a project undertaken to improve the observational network in the Guangdong province of China. The observing system is expected to provide high quality data for studying land-sea-air processes in the tropical monsoon area in China. The presentation identified some projects that can utilize the data from the observing system in conjunction with mesoscale models: studies on heterogeneous surfaces on the land-sea-air boundary layer processes in the tropical monsoon area in China; studies on the effects of the boundary layer processes on such destructive weather phenomenon as typhoons, heavy rains, and convection; and studies on the climatic effect of regional environment of haze, pollution, and atmospheric aerosols in the tropical monsoon area in China.

     Dr. Dev Niyogi gave the third presentation of Session II, entitled "Representing the Impacts of Urban Regions in Weather and Climate Models." Dr. Niyogi's presentation briefly introduced the different projects with which he is involved to identify collaborative research interests with the Chinese scientists. Topical areas include: water cycle and weather predictions, agricultural air quality and ozone deposition, crop and vegetation growth in climate models, aerosol and clouds/diffuse radiation, Carbon Cycle, droughts and terrestrial ecosystems in the southeast United States, and turf disease warning and forecasting. During the course of his presentation, Dr. Niyogi then gave an overview of land surface modeling, and the direct effects of an urban area, including heterogeneity induced dynamical effects (such as convection and heavy precipitation), model simulations in urban areas, and the need for representing the land surface processes in the vicinity of urban regions to improve the model simulations over urban areas. The presentation also highlighted the need for representing indirect effects of urban areas, such as pollution and aerosols and associated radiative feedbacks. Overall, the representation of urban effects in weather and climate land surface models is still in its infancy, but is rapidly evolving, and the scale issues need to be resolved, along with representation of indirect effects, which leads to possibilities for significant collaborative opportunities.

     Dr. Chaolin Zhang of IUM gave the final presentation of this session, and it was entitled "Numerical Experiments and the Impact of Land Surface Characteristics on Beijing's Summer Water Resource." Beijing has a significant water shortage problem. The changes in precipitation patterns and the increasing demand on water resources make investigation of the impacts of the weather on the availability of potable water a high priority. The study found very significant differences in the number of extreme precipitation events during the summer - one of the leading contributors to the annual potable water availability over the last few decades. They performed numerical studies and found that land surface characteristics can play a very important role in summer extreme rain events, and that adjustments to city planning and tree planting schemes can potentially impact rainfall distribution and the water supply for Beijing. They reported their findings that planting grass in select fringe regions around the urban centers could be sufficient for causing precipitation changes. The heterogeneity in surface forcings can lead to very complex interactions with rainfall, and that new observing and modeling techniques should contribute to obtaining better knowledge of these effects.

2.4 Session III: Urban Air Quality and Pollution

     Session III was co-chaired by Mr. Xiaochun Zhang of IUM and Dr. Kaihon "Alexis" Lau of the Hong Kong University of Science and Technology. Professor Jiayi Chen led the session off with a presentation entitled "The Mesoscale Flows and Air Pollutant Transport in Beijing." The objectives of the study were to study the climatology of the wind flows in the Beijing area (focusing on the lower part of the PBL) and providing meteorological information for air quality improvement planning. They found three general flow patterns that occur in the Beijing area: i) synoptic-scale flow showing stronger northerly wind in the winter and spring, and southerly or southeasterly in the summer; ii) local scale flow caused by the difference of thermal effects of mesoscale relief; and iii) miscellaneous - changing weather systems, stagnant air. They also found that the unfavorable diffusion regimes occur frequently as the local, diurnal circulation in Beijing is varying, and that the pollution potential is worst under stagnant weather conditions. The study also reported that emissions for the industrial district in the western part of Beijing typically impact the XiangShan Mountains and FenTai, rather than downtown Beijing, because of the typical flow regimes.

     Dr. Lau gave the second presentation of the session, entitled "Application of Urban Landuse Model in the Simulation of a Severe Air Pollution Event in the Pearl River Delta (PRD)." Dr. Lau reported that in the last ten years, the air quality in the PRD has declined dramatically, with the monthly number of days with visibility of < 8 km accompanied by relative humidities of < 80% (this combination should isolate haze days) increasing from upwards of 10 days per month in 1991 to around 25 days per month in the last few years. A series of land surface modeling analyses led to the conclusion that an understanding of local meteorological features is critical for forecasting the regional air quality. In the PRD region specifically, it was determined that while emissions have decreased in the Hong Kong area, they have risen as a whole throughout the PRD due largely to urbanization. The increased urbanization has also resulted in a stronger regional land-sea breeze circulation, which traps pollutants in the PRD during weak synoptic conditions.

     Dr. Guoan Ding of CAMS gave the next presentation, entitled "National Key Fundamental Research Project: 'Beijing Urban Environment Project' (Atmosphere Part)." The presentation detailed the Beijing City Air Pollution Observation Experiment (BECAPEX), which involved a 3-D observing network utilizing ground-based observations along with satellite datasets to monitor the urban environment in and around Beijing. During the course of the experiment, they were able to characterize the physical and chemical structure of the PBL surrounding Beijing, and depict various scales of BL phenomenon, such as a conceptual model of the urban air-dome, the urban heat island, etc. This group of researchers, led by Dr. Xiangde Xu (also of CAMS), also found that the air-dome acted like a large canopy covering Beijing, which was closely associated with the temperature inversion and vertical structure of the boundary layer, and that the diffusion and distribution of air pollutants were related to the multi-scale interaction among large-scale circulation, mountain-valley wind, and urban heat island circulations.

     Dr. Xiaoling Zhang of IUM gave the final presentation of the regular sessions. The presentation was entitled "The Monitoring Program and Primary Results in Shangdianzi Regional Atmospheric Background Monitoring Station." A study was undertaken to determine if Shangdianzi could serve as a regional background station. If it can, it is anticipated that Shangdianzi can provide long-term, accurate, and systematic data on the chemical composition and related physical characteristics and their trends in North China, and allow research to study the effects of upstream urbanization on downstream rural areas. This study confirmed that, in fact, Shangdianzi can be used as a good representation of the regional atmospheric background. Additionally, it was found that in general, the concentrations of air pollutants are lower in the background than in urban areas (except for ozone).

2.5 Session IV: Plenary Discussions

     The focus of this session was to summarize the proceedings of the first three sessions. All of the presenters for each session congregated to develop these summaries, which would be presented by one of the co-chairs of each session. The session presenters were asked to focus their summaries on the following:

• Session 1:
  • How should a UBL observing system be designed to meet the requirements for weather and climate research and applications in urban environments? Which sensors, procedures, and processes are needed?
  • Which measurement strategies will provide the best cost/benefit returns for operational decision makers?
  • What are the requirements across the spectrum of data gathering, collection assimilation, archival, and dissemination processes? How can we address them?
  • Are there new sensors and other measurement platforms available or emerging that could improve UBL observations?
  • What types of data sets are needed to improve UBL modeling?
• Session 2:
  • What meteorological information needed by urban decision makers is not yet available? Why?
  • How can we make better decisions, if given better meteorological information, for managing the safety, health, and economic well being of urban communities?
  • How could we quantify the socio-economic impacts of meteorological information on urban communities?
  • How can education, training, and outreach improve urban management and decision making with meteorological information?
• Session 3:
  • Do current meteorological observations meet the requirements of urban applications?
  • If not, in which areas are they lacking? What is the highest priority in terms of meteorological observations?
  • How can meteorological observations be improved to meet modeling and operational demand for urban areas?
  • How can we design an observation system for urban weather and climate monitoring?

     Dr. Sun presented a summary of Session I of the conference. She reiterated the problems and challenges that exist for obtaining measurements within each layer of the UBL. Additionally, she listed five types of measurements that urban models need in order to improve weather forecasts: i) surface roughness for momentum and scalar variables, ii) displacement heights, iii) boundary layer depths, iv) albedo and radiative properties of aerosols, and v) subsurface and urban canopy heat storage and water transportation properties.

     Dr. Niyogi gave Session II's summary. In this summary, he stated that presentations in this session showed that the addition of urban models has had positive results on the overall model performance and would assist numerical weather prediction. When dealing with urban modeling, typically the limiting factor on the complexity of the urban model is the availability of input data, which relates back to Session I's discussion on making observations within the UBL. Dr. Niyogi listed a few things that the presenters in the session felt were important to improve in order to advance urban modeling: i) estimation of latent heat flux in urban areas (many times this is assumed to be zero, which is incorrect); ii) update of landuse information, and some method of continuously (or at least on a monthly or seasonal basis) updating this information as development occurs; iii) consideration of indirect effects such as from aerosols, and iv) recognition that haze formation is a multifaceted problem, of which meteorology is only one aspect. The presenters felt that various research communities focusing on urbanization issues that are dedicated to specific regions (e.g. the Pearl River Delta region) are necessary and beneficial, because the urbanization problem appears to be very site-specific.

     Dr. Lau summarized the third session, during which he stated that the common theme of all the presentations during the session was that individual regions have specific phenomenon that are most important for that region's air quality. Many of the session's presentations were of some study that used some sort of observed air quality data. With this theme in mind, Dr. Lau turned to the issue of air quality data during the summary presentation. The group of presenters felt that the continual quality control of air quality measurements might be the most important factor in using air quality data. Further, they suggested that some sort of standard reporting for this quality control be developed, so that individual researchers have the ability to determine whether any set of data meets research grade or not. The importance of expanding the remote sensing of air quality data was also stressed. Finally, the session presenters stressed that work should be undertaken to better understand the feedback of air quality on meteorology, instead of focusing solely on the affect of the meteorology on air quality.

     The workshop co-chairs developed a summary of the presentation and collaborative opportunities possible between U.S. and Chinese meteorologists. Some of the collaborative ideas included several projects on: surface albedo, including evaluation of remotely-sensed albedo from MODIS, utilization of improved albedo over urban areas in NWP models, impacts of urbanization on albedo and the effect on the local/regional weather and ecosystems, and the installation of an albedo observational network; aerosol loading and air pollution problems, including the analysis of air pollution data, development of measurement techniques for diffuse radiation and evapotranspiration in urban areas, and air pollution modeling; the analysis and evaluation of the BUBBLEX 2004 dataset; the continued development of an urban model for Beijing; and the development of the real-time WRF/UCM forecasting capability for the support of the 2008 Summer Olympics in Beijing. It was also decided that a summary of the meeting's presentations and recommendations would be submitted for a peer-reviewed journal in China (in English).

3. CAMS Site Visit

     After the workshop, Dr. Niyogi and I visited the Chinese Academy of Meteorological Sciences (CAMS). While at CAMS, we were introduced to their Global/Regional Assimilation PrEdiction System (GRAPES) model, which is a global grid-point model. The GRAPES system is intended to be a multi-scale modeling effort, with application ranging from global ensemble studies to local weather forecasting. The dynamical parameterizations used in GRAPES are similar to those used in WRF, and much of the work done on one model may be ported to the other. Dr. Niyogi and Dr. Chen gave presentations including physical parameterizations related to land surface processes and its role on convective initiation and precipitation forecasting.

     In addition to the weather forecasting tasks, the CAMS is the only agency that develops a national weather forecast for television. We were able to watch the CAMS broadcast meteorologist create the national forecast presentation that would later be distributed to media outlets for TV broadcast after the noon news. CAMS produces four national forecasts a day, and all media outlets play the same national forecasts. We were also able to meet with the researchers working on remote sensing and radiation measurements networks and discussed potential collaborative opportunities related to extending the analysis on land-atmosphere interaction conducted over the U.S. to the southeast Asia region.

4. Conclusions

     Attendance at the Joint IUM/NCAR Workshop on Urban Meteorology: Observations and Modeling, and the opportunity to visit the CAMS, was a once-in-a-lifetime opportunity for me that I will always be thankful for. My conference experience is limited, but I was impressed with the professional manner in which this workshop was conducted, and the incredible amount of content contained within each presentation. I hope that the summaries contained in this paper even begin to give the presentations justice.