Grantee Research Project Results
2002 Progress Report: Implications of Climate Change for Regional Air Pollution, Health Effects and Energy Consumption Behavior
EPA Grant Number: R828731Title: Implications of Climate Change for Regional Air Pollution, Health Effects and Energy Consumption Behavior
Investigators: Ellis, Joseph H. , Samet, J. , Hobbs, Benjamin F. , Patz, J. F. , Joutz, Frederick L.
Current Investigators: Ellis, Joseph H. , Hobbs, Benjamin F. , Joutz, Frederick L.
Institution: The Johns Hopkins University
EPA Project Officer: Chung, Serena
Project Period: September 1, 2000 through August 31, 2003 (Extended to February 7, 2005)
Project Period Covered by this Report: September 1, 2001 through August 31, 2002
Project Amount: $1,376,739
RFA: Assessing the Consequences of Interactions between Human Activities and a Changing Climate (2000) RFA Text | Recipients Lists
Research Category: Climate Change , Air
Objective:
This research program has four major modeling elements: climate change and variability, electrical energy demand and production, regional air pollution, and human health effects associated with air pollution exposure. Our overall objective is to develop a scientifically credible modeling facility that will help policymakers and analysts understand the effects of human activities on climate change and variability as well as the possible human responses and adaptations to climate change and variability. The overall connections among the four modeling elements are shown schematically below.
Progress Summary:
Progress to date is described below for the three major modeling efforts underway—Electric Power Emission Impacts, Hourly Electricity Load Modeling and Forecasting, and Regional Air Pollution Modeling and Epidemiological Analyses.
Electric Power Emission Impacts – B.F. Hobbs, Y. Chen. The objective of Year 2 of the project's tasks in the electricity simulation area was to improve our first-year estimations of climate change impact on emission from the electricity sector. The following tasks were undertaken in pursuit of this objective: (1) we incorporated a transmission network representation in our model, which allows us to more accurately predict spatial emission variations across states in the Pennsylvania-New Jersey-Maryland (PJM) region (see Figure 1); (2) we included the U.S. Environmental Protection Agency (EPA) Ozone Transport Commission's (OTC) NOx Budget Program (cap-and-trade) in our model, which enables us to understand the interaction of electricity and nitrogen oxides (NOx) permit market, and to account for the effect of limited supplies of emissions allowances on emissions under alternative climate scenarios; and (3) we established a link with the Load Forecasting Group at George Washington University, permitting us to perform analyses of the effect of ambient temperature changes on emissions during selected ozone episodes.
Figure 1. The Overall Connections Between the Four Modeling Elements
In summary, given our assumptions regarding climate change, we performed an emissions analysis for each of three scenarios concerning the scope of the OTC NOx Budget Program: 2000 cap, 2001 cap, and full cap. The full cap covers all NOx sources, while the other scenarios represent the actual (partial) coverage by the OTC. The results show that a 2°F increase in ambient temperature translates into a 4.9 percent increase in load, and a 5 percent increase in total NOx emissions in the 2000 cap case. However, the impact on emissions during the warmest 3-day period in the study timeframe was approximately twice that, indicating that effects on ambient ozone and health likely will be greater during that time than on average during the entire ozone season. The NOx increases are divided into two effects: a load effect (emissions increases resulting from the increase in power demanded) and an efficiency effect (emissions increases resulting from temperature-induced decreases in power-plant fuel efficiency). The load effect is found to have roughly the same magnitude of impact as efficiency effect. The total effect is approximately equal to the sum of load and efficiency effect.
With the inclusion of the transmission network, our analysis shows that impact on state emissions levels is quite divergent. We report a significant impact of 25 percent increase in Delaware NOx emissions and less than 10 percent increase in Maryland and in Pennsylvania. However, New Jersey reduces its emissions by 15 percent in the first two scenarios, in contrast to an 8 percent increase in the full cap scenario.
Analysis of the time distribution of emissions shows that the impact of climate-warming on emissions is most profound during peak-load hours and less in off-peak hours. To the extent that total emissions are limited by the emissions cap, climate warming results in a shifting of emissions from cooler days to warmer days, which are more likely to coincide with ozone episodes.
Hourly Electricity Load Modeling and Forecasting – F.L. Joutz, C. Crowley. The objective of Year 2 of the project’s tasks in the electricity demand area was to develop hourly electricity load models that can be used to test for the effects of temperature and climate variability. The following tasks were undertaken in pursuit of this objective: (1) we constructed databases of hourly weather and electric load observations for all 10 load regions in the PJM Interconnect; (2) in response to comments from presentations we enhanced the hourly temperature sensitivity (quadratic) models by including cubic specifications; (3) we estimated and tested models for all 10 load regions; (4) we derived estimates of temperature elasticity of demand; (5) we began long-run forecasting and simulation research; and (6) we coordinated with the JHU team in linking output from temperature load sensitivity models with the network generation and transmission network models.
With regard to temperature sensitivities, we found: (1) nighttime elasticities (between 10:00 p.m. and 5:00 a.m.) are significant and measure around 0.5 percent-1.0 percent. This probably reflects the demand for comfort while sleeping even though this is the trough load period. Air conditioning demand is responsive on very hot nights; (2) in the morning, temperatures and loads rise as people and businesses start their daily activities. From 7:00 a.m. to 11:00 a.m., the elasticity averages about 1.5 percent; (3) at mid-day, temperature and load approach their daily peaks. The load sensitivity is marginal from 11:00 a.m. to 2:00 p.m.; and (4) from 2:00 p.m. through 6:00 p.m., the elasticity is greater than 1 percent. Between 6:00 p.m. and 9:00 p.m., the elasticity appears to be zero. This counter-intuitive result is probably explained by offsetting effects: at the close of the business day, residential demand for electricity increases while businesses are closing. In addition, the temperature may by high, but it is falling during the evening hours.
Regional Air Pollution Modeling and Epidemiological Analyses – M. L. Bell, J.H. Ellis, J. Patz and J. Samet. In Year 2 of the project, we accomplished four major modeling tasks: (1) sensitivity analyses using our existing Models-3 framework; specifically, assessing the effects of changes in biogenic and NOx emissions on subsequent tropospheric ozone and p.m. ambient concentrations; (2) examination of the effect of the new ozone National Ambient Air Quality Standards (NAAQS) on exceedences in Baltimore and Atlanta; (3) exploratory analyses mapping changes in ambient ozone concentrations to selected public health impacts, using a wide variety of published epidemiological models drawn from the EPA Section 812 report, and using a suite of modified emissions involving biogenics and mobile source NOx; and (4) implementing and extensively testing an updated regional air pollution modeling system comprised of Mesoscale Model 5 (MM5) (Version 3.6.1), Meteorology Chemistry Interface Processor (MCIP) (Version 2.1), Sparse Matrix Operator Kernal Emissions (SMOKE) (Version 1.4), and Models-3 (Community Multiscale Air Quality [CMAQ] 4.2.2).
Task 1. Sensitivity analysis was performed using an emissions scenario with a 100-percent increase in biogenic emissions to explore how such increases could impact ozone concentrations for a case study episode. Another emissions scenario using an additional 100-percent increase in motor vehicle emissions of ozone precursors allowed comparison of the effects of increased biogenic volatile organic compounds (VOCs) to that of an anthropogenic source.
Biogenic VOCs had a greater impact than comparable percent increases in motor vehicle emissions of ozone precursors. This corresponds to the high fraction of VOCs that are of biogenic origin. The 100-percent increase in biogenic VOC emissions raised ozone levels, with an estimated maximum 1-hour concentration that was 30 percent higher than that of the baseline scenario. The additional emissions of ozone precursors from motor vehicles raised the maximum 1-hour concentration 40 percent more than that of the baseline. The largest increases in ozone concentrations occurred near peak values. Urban areas had larger increases in ozone levels than rural regions. Both adjusted emissions scenarios resulted in ozone concentrations lower than that of the baseline for some estimates; however, these reductions occurred near low ozone levels and generally were small. These results clearly demonstrate the significance of biogenic VOC emissions in ozone formation for this region and emphasize the difficulty in controlling ozone levels and the importance of biogenic emissions inventories. These results also imply that climate-change induced increased biogenic VOC emissions could significantly impact ozone concentrations.
Task 2. In 1997, the EPA introduced a revised NAAQS governing tropospheric ozone, adding an 8-hour standard of 0.08 ppm and phasing out the 1-hour requirement of 0.12 ppm. The 8-hour standard is intended to provide greater protection for human health. Here, we examined the spatial and temporal patterns of exceedances of the 1-hour and 8-hour ozone NAAQS using monitoring data and ozone concentration estimates. MM5 Version 3-4 was used to generate estimates of meteorological variables. The Models-3 framework was used to estimate hourly ozone concentrations for each of 2,700 4-km by 4-km gridcells for a MD/VA/DE/DC domain and 2,880 gridcells for a northern Georgia domain. Hourly ozone measurements from 29 monitors were obtained from the EPA Aerometric Information Retrieval System (AIRS) and state environmental agencies. Monitor measurements were analyzed for the MD/VA/DE/DC case studies and for all monitors in Maryland for 1995. The population living in areas with NAAQS exceedances was estimated. Our results show that the spatial and temporal nature of compliance with the ozone NAAQS is different under the 8-hour standard. All results indicate that the revised standard is exceeded more often and in more places than the original 1-hour requirement. In the modeling simulations, the 8-hour standard was exceeded 2.0 to 5.2 times more often than the 1-hour standard, and was exceeded in 1.8 to 16.2 times more area. Results further reveal that the 8-hour standard is exceeded in areas that generally comply with the 1-hour standard and are not well covered by the monitoring network. This implies that the current monitoring system, which focuses on urban and suburban areas, is not sufficient to detect exceedances of the 8-hour standard. In addition, these results imply that a larger population resides in areas with unhealthy ozone levels than noncompliance with the original 1-hour standard suggests. For the MD/VA/DE/DC domains, approximately 6.5 million and 8 million people reside in areas with 8-hour NAAQS exceedances for the 1990 and 1995 episodes, respectively. This constitutes 80 to 98 percent of the domain’s total population.
Task 3. The epidemiological analyses in this research project estimated the public health impacts of six scenarios with different emissions of ozone precursors. Epidemiological studies were used to generate concentration-response functions to assess changes in health outcomes corresponding to the ozone levels for the emissions scenarios. Total and cause-specific mortality, hospital admissions, and emergency room visits were examined. The largest increases in ozone levels, and therefore health impacts, resulted from elevated levels of biogenic VOCs. A 25-percent increase in anthropogenic and biogenic VOCs and NOx resulted in an estimated increase in mortality up to 0.72 percent, when averaged across the spatial domain. Results indicate that the health impacts of altered emissions would not be uniformly distributed. Large increases in ozone levels, and therefore health effects, were estimated for western Pennsylvania, the Virginia coast, the District of Columbia, and the area east of Baltimore, MD. Even emissions strategies that resulted in an overall decrease in ozone levels, such as would arise from a 25 percent increase in anthropogenic NOx and for this case study, would cause detrimental health impacts from higher ozone concentrations in some areas. In this scenario, the average daily mortality associated with ozone would decrease by less than one percent; however, some regions would experience an increase of approximately 3.8, 1.7, 1.7, and 1.7 percent in daily total, respiratory, cardiovascular, and chronic obstructive pulmonary disease (COPD) mortality, respectively. These results are based on midpoint estimates calculated using concentration-response functions from epidemiological studies. In areas with the most significantly elevated ozone levels, a 25-percent increase in anthropogenic and biogenic VOCs corresponded to an increase in hospital admissions for the elderly of 12, 8.5, and 9.0 percent for total respiratory, pneumonia, and COPD causes. Pediatric asthma emergency department visits would rise as much as 2.7 percent in this scenario. The largest increase in daily mortality for any region was a 5.1-percent increase in the scenario that increased all VOC emissions by 25 percent. Results indicate that increases in ozone precursors could cause serious human health consequences through higher ozone levels, and that the impacts would not be distributed uniformly from a spatial perspective. This research isolates the health consequences from tropospheric ozone levels corresponding to multiple emissions scenarios. This does not consider the health impacts of VOCs and NOx that are independent of their contribution to ozone formation. Furthermore, not all health effects of ozone were examined, such as respiratory symptoms and asthma attacks. Due to the serious and widespread nature of ozone health impacts, these increases in mortality, hospital admissions, and emergency room visit rates represent an important health concern and challenge for decisionmakers and those designing tropospheric ozone control strategies.
Task 4. The air pollution modeling system described above differs substantially from the prior system in that the original Models-3 framework had been completely abandoned, and SMOKE is used for emissions processing, instead of MEPPS. An important consequence of this change is that the system now is free of 2 GB maximum file size limitations (all model runs now are performed daily; virtually any length simulation period can be modeled). Also, the new system uses gridded spatial surrogates (used in biogenic and area source emissions preparation) from the "so-called" U.S. Unified Grid, thus obviating the need for SAS and ARC/Info in the system. This new system was implemented on both the existing Sun Ultra-60/Solaris platforms and a recently acquired Dell Precision Workstation running Linux and using Portland Group Fortran and C compilers (the PC-based system is remarkably faster). We require further testing regarding the credibility of the gridded spatial surrogates used in the generation of biogenic and area-source emissions. We have noted some discrepancies between prior analyses (using the Models-3 framework and MEPPS) and the current system with regard to the emissions fields that are subsequently input to the photochemical modeling routines (CMAQ). On occasion, we find that the new system seems to significantly underpredict some emission types (e.g., NOx) with the expected outcome that the resulting ozone concentrations exhibit a restricted dynamic range (not low nor high enough, when compared with measured data). The next set of analyses will employ the well-tested U.S. 132x90 36-km grid and its gridded surrogates, using meteorologic input generated inhouse (a month-long episode in August 1996). If these results show reasonable ozone dynamic range, then the unified grid surrogates must be somehow corrupt or otherwise nonrepresentative.
Future Activities:
The future activities are listed below by investigators.
Drs. Hobbs and Chen plan to:
· Expand the study region to include Kentucky, Virginia, Ohio, West Virginia, and eastern Indiana, whose NOx emissions also significantly contribute to ambient ozone in the mid-Atlantic states.
· Develop a long-run market model for capacity mix to simulate responses of the structure of the power generation system to climate warming-induced changes in demand and generator characteristics.
· Link the emissions and air pollutant transport and transformation models to evaluate ambient air quality impacts of the emissions changes.
· Incorporate other air pollution related policies such as Clear Skies Initiative and Renewable Portfolio Standard in model.
Drs. Joutz and Crowley plan to:
· Include further load sensitivities for the Supply Dispatch and Generation Modeling Group at JHU.
· Consider alternative scenarios, including winter effects.
· Develop longer-term sectoral electricity demand models (e.g., use National Energy Modeling System, as developed by Energy Information Administration, and consider Residential End-use Energy Planning System (REEPS) and COMMercial sector END-user planning system (COMMEND), as developed by the Electric Power Research Institute.
· Incorporate price, income, and technological effects in addition to climate change and variability.
Dr. Ellis plans to conduct tests of the credibility of the gridded spatial surrogates used in the generation of biogenic and area source emissions. We have noted some discrepancies between prior analyses (using the Models-3 framework and MEPPS) and the current system with regard to the emissions fields that subsequently are input to CMAQ. On occasion, we find that the new system seems to significantly underpredict some emission types (e.g., NO) with the expected outcome that the resulting ozone concentrations exhibit a restricted dynamic range (not low nor high enough when compared with measured data). The next set of analyses that I will perform will employ the well-tested U.S. 132x90 36-km grid and its gridded surrogates, using meteorologic input generated in house (a month-long episode in August 1996). If these results show reasonable ozone dynamic range, then the unified grid surrogates must be somehow corrupt or otherwise nonrepresentative.
Execution of regional air pollution simulations using modified emissions and climatic scenarios driven by the results of Drs. Hobbs, Chen, Joutz, and Crowley.
Drs. Patz and Samet plan to use the meta-analysis of ozone dose-response (conducted by Michelle Bell and described above) for future ozone-related epidemiological analyses. The Department of Epidemiology also identified the following sources to review for particulates: Section 812 Report Appendix, Particulate Criteria Document, Centralia Power Plant Risk Assessment, ABT report on quantifying the benefits of reducing power plant emissions, and St. George's Hospital Medical School’s database. For these studies, a check for potential bias due to the recent GAM problem identified in the S-Plus default will be evaluated.
Regarding baseline mortality data for the five-state study area, we most likely will use the NMMAPS database (within the departments of Epidemiology and Biostatistics) that has baseline mortality data from 1987-1994, and most likely now- 1999, will be used to compare future air-pollution related mortality. By focusing our analysis on metropolitan areas included in the NMMAPS studies, we might then bypass the need for fitting census data to model grid cells (we have recently acquired a Medicaid database, and in our next meeting we will decide if this database would be advantageous to use for this project). Also, we have conducted an updated literature review on climate/air pollution/health.
Journal Articles on this Report : 5 Displayed | Download in RIS Format
Other project views: | All 24 publications | 10 publications in selected types | All 9 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Bell M, Ellis H. Comparison of the 1-hr and 8-hr National Ambient Air Quality Standards for ozone using Models-3. Journal of the Air & Waste Management Association 2003;53(12):1531-1540. |
R828731 (2001) R828731 (2002) R828731 (Final) |
Exit Exit |
|
Bell M, Ellis H. Sensitivity analysis of tropospheric ozone to modified biogenic emissions for the Mid-Atlantic region. Atmospheric Environment 2004;38(13):1879-1889. |
R828731 (2001) R828731 (2002) R828731 (Final) |
Exit Exit Exit |
|
Chen Y, Hobbs BF. An oligopolistic power market model with tradable NOx permits. IEEE Transactions on Power Systems 2005;20(1):119-129. |
R828731 (2002) R828731 (Final) R828733 (Final) R831836 (2005) |
Exit Exit |
|
Joutz F, Crowley C. Seasonality and weather effects on electricity loads: modeling and forecasting. Energy Policy. |
R828731 (2002) |
not available |
|
Munson T, Leyffer S, Chen Y, Hobbs B. Comparisons of MPEC algorithms for a leader-follower market equilibrium problem: electric power and NOx allowance markets. Computational Optimization and Applications. |
R828731 (2001) R828731 (2002) |
not available |
Supplemental Keywords:
regional air pollution, electricity demand forecasting, electrical system dispatch, health effects, ozone, particulate matter, PM, climate change, air toxics, troposheric ozone, PM2.5, PM10, ambient air pollution, climate models, climate variability, climate variations, ecosystem sustainability, electrical energy, emissions inventory, energy generation, exposure and effects, human activity, human exposure, integrated assessments, policymaking, Delaware, DE, Maryland, MD, Virginia, VA, Washington, DC., RFA, Health, Scientific Discipline, Air, Geographic Area, particulate matter, air toxics, Health Risk Assessment, climate change, State, Risk Assessments, tropospheric ozone, Atmospheric Sciences, integrated assessments, electrical energy, PM10, environmental monitoring, exposure and effects, stratospheric ozone, policy making, Virginia (VA), Delaware (DE), human activities, PM 2.5, energy generation, climate variations, climate models, Maryland (MD), emissions inventory, human exposure, DC, PM, ecosystem sustainability, human activity, climate variability, ambient air pollution, Global Climate ChangeProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.