Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
42 Integrated Modeling and Measurement Recommendations 5.1 Introduction Although from a resource standpoint, modeling is usually preferred over measurements, the need for integration of both may be required largely because of modeling limitations. Even with all of the advances that have been made in developing and improving models to sup- port airport air quality assessments, there still remain significant modeling gaps (e.g., using AERMOD) that are either difficult or impossible to fill at this time. Measurements are partly necessary because chemically reactive species and secondary pollutants such as ozone cannot be easily and robustly modeled. Although CMAQ helps to fill several of these modeling gaps, there are issues associated with fidelity and spatial resolution that need to be addressed. Also, measurements generally provide more accurate and more defensible results than modeling, but lack adequate characterization of spatial and temporal variability and can be resource-intensive (expensive). The next two subsections provide a general, decision-making framework that can be applied to most airport scenarios. However, this framework needs to be integrated with the specific needs of each airport as well as the regulatory context of the air quality assessments. For example, the goals under a NEPA project may be different than those under a local (e.g., city) requirement. 5.2 General Decision Process The decision on what models to use and measurements to conduct needs to be based on the specific needs of each project. No single combination of modeling and measurements will suit all airport projects (i.e., one size does not fit all). However, Figure 5-1 presents a general flow chart that can help to methodically determine the modeling and measurement combination to use, including which models. Once the combination is determined, the guidance from Chapters 3 and 4 should be followed to implement any modeling or measurement programs. Figure 5-1 represents one decision-making order. Others are possible by ordering the decision points differently, but the outcome would be the same for each scenario. Therefore, the order is not as important as the need to make sure each of the decision points (as reiterated below) is considered: ⢠Assess stable pollutants (e.g., CO, PM2.5, PM10, NO2 and SO2)? ⢠Need to support detailed assessments, including troublesome hot spots? ⢠Need to establish accurate baselines and/or meet local (e.g., state) requirements for measurements? ⢠Regional concentrations necessary? C h a p t e r 5
Integrated Modeling and Measurement recommendations 43 ⢠Need future year concentrations? ⢠Resources available? ⢠Reactive or secondary included in CMAQ? ⢠CMAQ 4 km by 4 km resolution sufficient? The starting point for the decision-making process in Figure 5-1 is the determination of which pollutants to assess. In general, EDMS/AERMOD is used to model relatively stable pollutants (e.g., CO, PM10, PM2.5, NO2, and SO2) while CMAQ and measurements can be conducted to assess reactive pollutants. As previously indicated, some reactive pollutants can be modeled in EDMS/AERMOD, if the reaction rate is relatively low (i.e., lower than the duration required for the pollutant to travel from the source to the farthest modeled receptor location). Although this decision point results in the first set of major decision branches, they should not be considered mutually exclusive. Indeed, the ensuing sub-branches merge as necessary. If both stable and unstable (relatively reactive) pollutants are to be assessed, then the flow chart can still be used, especially to help determine the use of measurements versus CMAQ. If the pollutants are relatively stable, then the next decision point concerns the potential need to support detailed and troublesome assessments. Although models can be used to conduct such work including hot spot assessments, there may be some specific reasons (e.g., features at an airport) that may make it difficult to model certain locations and/or the locations may have Figure 5-1. Modeling and measurements decision support diagram.
44 Guidance for Quantifying the Contribution of airport emissions to Local air Quality significant community health concerns. In such cases, measurements may be necessary to either help confirm the modeled results or provide more accurate concentrations. Another reason for measurements may be due to either an airport initiative to better under- stand its impacts on the local community or in response to local government requirements. In any case, a set of baseline measurements (e.g., a yearsâ or multiple yearsâ worth of concentration mea- surements) may be necessary to determine average levels as well as impacts from meteorology. In addition to the local-scale impacts for relatively stable pollutants that can be assessed using EDMS/AERMOD, the larger scale modeling capabilities of CMAQ may be used to help deter- mine regional impacts. As such, the need for larger scale results needs to be weighed with the resource requirements needed to use CMAQ. In assessing both reactive and secondary pollutants, one of the decision points that will need to be addressed is whether future year concentrations will need to be predicted. Associated with this is whether or not air quality for varying scenarios (e.g., âwhat-ifâ-type modifications to aircraft operations and taxi movements) will need to be determined. Although measurements may provide accurate concentration levels (assuming sufficient samplings), they generally do not allow assessment of different scenarios. Another decision point regarding reactive pollutants is to make sure CMAQ can ade- quately model their concentrations. A thorough model evaluation needs to be conducted (as demonstrated in this project), and the evaluation needs to be compared with similar studies to see if the measures of model performance are comparable or within the criteria and goals recommended by the EPA. As such, expert decisions will need to be made whether CMAQ predictions will suffice or if measurements will need to be conducted either in support of or as a replacement to the use of CMAQ. The other decision regarding the use of CMAQ to model reactive pollutants involves the resolution of the grids used. Although smaller grids may be possible, the smallest that has been used to produce acceptable results is 4-km grids (16 km2 areas). This means that concentrations within each 4 km by 4 km grid are homogenous (i.e., one average concentration for each grid). If higher resolution concentrations are necessary, then either higher resolution grids will need to be investigated and/or measurements will be necessary. A common decision point when considering measurements is whether adequate resources (e.g., personnel, equipment, and time) are available to conduct the work. Depending on the scope of the work, it can often be difficult to determine the overall resources necessary to conduct measurements. Hence, underestimating resources is not uncommon when con- ducting measurements, and this may be compounded by unforeseen (or at least difficult to foresee) events such as inclement weather. Although resources are only highlighted for measurement work in Figure 5-1, it should be obvious that the various modeling work require resources as well. Measurements and modeling each require personnel expertise (e.g., airport personnel and consultants) to conduct both types of work and should not be overlooked. Therefore, the issue of resources should actually be considered throughout the decision-making process. While all of these decision points provide a general framework that can be applied to each airport, they need to be tailored to meet the needs of each specific airport. Some of the needs and specific issues at airports may include but are not limited to ⢠Community and political pressures to use certain types of resources; ⢠Pressures to conduct measurements rather than modeling; ⢠Availability of appropriate background concentrations (e.g., from USEPA monitoring stations);
Integrated Modeling and Measurement recommendations 45 ⢠Local regulatory requirements; ⢠Desire to conduct the modeling work only using airport personnel; and ⢠Little confidence in the models. Each of these types of issues will need to be considered as the aforementioned decision points are addressed. A specific issue may help to confirm the decisions made, contradict and override the decisions, or, depending on the scope of the issue, may not impact certain decisions. 5.3 Modeling Versus Measurements To better understand the decision points, it is also important to review the advantages and dis- advantages of using models and conducting measurements. Although some of these points were covered in the previous section, they are expanded and elaborated on in this section to provide a better understanding of the issues involved. In general, the reasons for modeling tend to be dis- advantages for conducting measurements and vice versa. Some of the advantages of modeling include but are not limited to ⢠Generally more cost-effective than measurements, ⢠More flexibility in spatial and temporal coverage, ⢠Can predict future concentrations, and ⢠Allows âwhat-ifâ-type scenario modeling. In general, modeling is less costly than measurements. However, this will depend on the over- all scope of the assessments as it is conceivable that some modeling work could be comparable or even more expensive. The overall modeling work includes all of the tedious work required to gather the various input data. There are various degrees of fidelity that can be achieved in the resulting concentrations based on the quality of the input data. Therefore, it is important that the model users understand the impacts of each set of input data. By understanding these impacts (sensitivities to the input data including the meteorology and locations of receptors), modelers can optimize their use of resources by focusing their efforts on data that have the most impact. But even with all of these efforts and considerations that go into modeling, measurements are usually much more expensive. All of the planning, scoping, sample collections, and so forth associated with measurements can easily dwarf the costs of modeling. With regard to spatial and temporal coverage, models usually have a significant advantage in providing the ability to model various receptor locations and, depending on the availability of appropriate source activity and fleet data, different times can be modeled. Even if resources were available to set up a wide array of field sampling locations, the logistics including safety and other airport restrictions may not allow such setups. And unless real-time analyzers are used, sampling at fine resolutions (e.g., every hour of a day) may not be possible. As such, modeling allows for the flexibility to predict concentrations at fine spatial and temporal reso- lutions that could be valuable in conducting detailed assessments to better understand trends and airport impacts. However, it should be pointed out that while this is true for Gaussian models such as AERMOD, it is not wholly true for models such as CMAQ. Although CMAQ provides the ability to model chemical transformations of various pollutants, including sec- ondary species, the spatial resolution is typically limited to the grid sizes (e.g., 4 km by 4 km) used for modeling. Most NEPA projects that require air quality dispersion modeling will need the ability to predict concentrations corresponding to future scenarios, including both build and no-build cases. While this is related to the flexibility issue, it illustrates the inherent limitation of mea- surements, which can only be conducted for current conditions. Of course, predictions of future
46 Guidance for Quantifying the Contribution of airport emissions to Local air Quality concentrations need to be carefully considered as their accuracy is dependent on the quality of the projected input data and any assumptions thereof. Similar to predicting future concentrations, the ability to play âwhat-ifâ games also pro- vides models with a significant and powerful advantage over measurements. While some scenario modeling is certainly possible, they are usually dependent on either the opportuni- ties that present themselves through the natural operation (differences from hour to hour or day to day) of an airport or through staged operations (scenarios created by the airport) provided by the airport. In either case, the opportunities to analyze different scenarios would be limited by what is possible at an airport (e.g., use of different aircraft and GSEs). In con- trast, various policy scenarios can be modeled and this is only limited by the quality of the available input data. As these discussions point out, modeling can provide many advantages related to costs and robustness with regards to scenario modeling. While these are all disadvantages to conducting mea- surements, there are various advantages as well that can be pointed out including but not limited to ⢠Develop more accurate yearly concentrations; ⢠Spatial resolution of chemical transformation models (but can be limited depending on resources); ⢠Clarify difficult modeling assessments; ⢠Develop more accurate background concentrations; and ⢠Obtain more local meteorological data. Either due to community pressures or local government requirements, an airport may want to develop a measurement program that monitors air quality in and around the airport. This may be a continuous program that monitors air quality yearly or for a finite period. In any case, the measured results represent a level of accuracy results that can be used to assess airport contributions to local air quality. Notwithstanding any limitations from the amount of pol- lutants covered or spatial and temporal resolutions, a long-term measurement program could provide valuable data that could serve as accurate baselines for comparison purposes and allow trend assessments. The main point is that measured data are considered more accurate and more defensibleâsuch data constitute the âgold standard.â Furthermore, the usefulness of measured data increases when continuous monitoring equipment is used to obtain highly time-resolved data as opposed to sampling equipment that requires laboratory analysis. As previously indicated, the current grid models such as CMAQ that can be used to predict concentrations of reactive species are limited by the resolution of their grids. A single grid could include an entire airport and its surrounding community. While CMAQ could provide relatively accurate concentrations within a region containing an airport, it could be difficult to obtain dif- ferences in concentrations within the region. Depending on the size of the region, placement and different sized grids could be used. Although the 4 km by 4 km grid size is currently the smallest size that has been used and tested with CMAQ, smaller sizes may be possible and are being explored in ongoing research using adaptive or variable-grid resolution approaches. However, hybrid model- ing approaches such as that used in this study using CMAQ and AERMOD provide a powerful alternative to developing explicit very fine-scale model applications using CMAQ alone. Although models can be robust in providing a wide variety of capabilities that measurements cannot, measurements can still ultimately provide the ârealâ or âactualâ information that can help address uncertainties in modeled results. Notwithstanding the representativeness (spatially and temporally) of the measured results, they can be used to fill modeling deficiencies. Such defi- ciencies may range from difficult-to-model chemical transformations to unusual geographical features.
Integrated Modeling and Measurement recommendations 47 Because of the relatively low airport contributions of some pollutants (e.g., CO), it would be advantageous to develop better background concentrations through measurements. Obviously, such efforts would need to be weighed based on the need for increased fidelity versus availability of resources to conduct the measurement work. In addition to data obtained from various existing meteorological stations (e.g., NOAA Auto- mated Surface Observing Systems [ASOS]), measured weather data from equipment located strategically in and around the airport could provide a better understanding of the wind fields that affect the movement and dispersion of pollutants. Such data would be beneficial to any assessments involving the use of measured concentrations but could also be used to improve modelingâe.g., better modeling of future scenarios.
