![[logo] US EPA](https://webarchive.library.unt.edu/eot2008/20081105141143im_/http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
2002 Progress Report: Detection of a Recovery in Stratospheric and Total Ozone
EPA Grant Number: R829402C001Subproject: this is subproject number 001 , established and managed by the Center Director under grant R829402
(EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).
Center: Center for Integrating Statistical and Environmental Science
Center Director: Stein, Michael
Title: Detection of a Recovery in Stratospheric and Total Ozone
Investigators: Tiao, George , Meng, Xiao-Li , Reinsel, Gregory , Weatherhead, Betsy , Wuebbles, Donald J.
Current Investigators: Tiao, George , Fioletov, Vitali , Flynn, Lawrence , Guillas, Serge , Hayhoe, Katharine , Kerr, James , Meng, Xiao-Li , Miller, Alvin , Petropavlovskikh, Irina , Reinsel, Gregory , Weatherhead, Elizabeth , Wuebbles, Donald J. , Yang, Shi-Keng
Institution:
Current Institution: Environment Canada , Harvard University , National Oceanic and Atmospheric Administration , University of Chicago , University of Colorado , University of Illinois at Urbana-Champaign , University of Wisconsin - Madison
EPA Project Officer: Smith, Bernice
Project Period: March 12, 2002 through March 11, 2007
Project Period Covered by this Report: March 12, 2002 through March 11, 2003
RFA: Environmental Statistics Center (2001)
Research Category: Environmental Statistics , Ecological Indicators/Assessment/Restoration
Description:
Objective:The Center for Integrating Statistical and Environmental Science supports research in the development of new statistical methods for addressing environmental problems and for integrating the use of statistics throughout the process of risk assessment. The Center's main objective is to advance the use of statistical methods to assess the state of the physical environment and its impact on human and ecological health. This research investigates the recovery in stratospheric and total ozone using various modeling techniques. Work has been divided into five subprojects: (1) analysis of ground station total ozone data for detection of trend turnaround and recovery; (2) analysis of satellite total ozone data for detection of trend turnaround and recovery; (3) trend analysis of upper stratospheric ozone data for evidence of turnaround; (4) statistical diagnostic analysis of numerical models; and (5) combining information from statistical analysis and numerical models.
Progress Summary:Analysis of Ground Station Total Ozone Data for Detection of Trend Turnaround and Recovery
We have performed a statistical trend analysis of atmospheric total ozone data from a network of 36 ground stations in the northern mid-latitude regions of 25-60N for the period 1971 through 2002 for detection of a "turnaround" in the previous downward trend, and hence evidence for the beginning of an ozone recovery. Because other climatic and geophysical changes can impact on ozone behavior, and hence influence the detection of recovery, we include in the statistical-trend modeling and analysis the effects on ozone variability associated with various dynamical and circulational variations in the atmosphere (such as quasi-biennial oscillation [QBO] and arctic oscillation [AO] influences), influences of the solar cycle, and stratospheric aerosols from volcanic eruptions such as Mt. Pinatubo in the early 1990s.
Analysis of Satellite Total Ozone Data for Detection of Trend Turnaround and Recovery
The ground station analyses have been extended to study 5-degree zonal average time series from a total ozone data set of merged total ozone mapping spectrometer (TOMS) and solar backscatter ultraviolet (SBUV) satellite sources, as constructed by Stolarski and Hollandsworth Firth, and to study 10-degree zonal average time series from a cohesive total ozone data set of SBUV(/2) satellite instruments, as detailed by Miller, et al. (2002), for the Northern Hemisphere over the period November 1978 through December 2002. The main goal is to assess the degree of consistency of trend findings between ground station and satellite data.
Trend Analysis of Upper Stratospheric Ozone Data for Evidence of Turnaround
A similar approach also has been employed to examine upper stratospheric (~35-45 km) Umkehr ozone data over the period 1977-2001 from three high-quality stations—Arosa (47N), Boulder (40N), and Tateno (36N)—for evidence of turnaround in ozone trend behavior at this altitude region that is of critical importance in understanding the overall ozone trend behavior.
Statistical Diagnostic Analysis of Numerical Models
We have begun to develop an iterative method that combines information from atmospheric model ozone calculations with observed monitoring ozone data to assess the capabilities of the atmospheric model under study, pinpoint its possible deficiencies, and suggest directions for improvement. The basic idea is very simple. First, calculated ozone from the model is taken as input to an empirical statistical model for observed ozone data, for example, a simple regression model. The fitted models then are studied diagnostically to determine whether the atmospheric model adequately has taken into account the impact of exogenous factors such as solar cycles, QBO and volcanic aerosols, and whether seasonal variations and trend changes in the model calculations are consistent with those in the observational monitoring data. Information on the deficiencies then can be utilized by the atmospheric modeler to modify the model and produce a new set of ozone calculations for further comparison with the observed data, and the cycle is repeated.
Combining Information From Statistical Analysis and Numerical Models
The statistical diagnostic analysis above can be extended into a general approach that combines statistical evaluation of ozone changes and the scientific understanding of stratospheric ozone from numerical models. Once a numerical model, such as the University of Illinois–Urbana-Champaign (UIUC) two-dimensional (2D) model of the troposphere and stratosphere is judged adequate, outputs from the model can be used as inputs or predictors into a statistical model for observed ozone data in a variety of ways:
• As an input function for past trends in ozone to replace the traditional use of "hockey stick" trends.
• As a function for studying the recovery of stratospheric ozone, particularly in terms of the statistical interpretation of the ability of atmospheric observations to detect the recovery of ozone.
We believe that this approach is likely to achieve a synergy of information from observational data and numerical models that potentially can help shorten the time needed to determine whether or not the ozone recovery has been initiated.
Results to Date
Analysis of Ground Station Total Ozone Data for Detection of Trend Turnaround and Recovery. Our models for the trend analyses include terms for the linear trend from 1971, a change in linear trend assumed to occur in 1996 (suggested by current atmospheric modeling theory and casual empirical examination of the data), f10.7 cm solar flux for solar cycle effects, 50 mb zonal winds at Singapore (with time lag) for QBO effects, and an arctic oscillation index series (with time lags) for AO effects. As a sensitivity study, we also estimate the trends without the AO index series in the model. A notable result is that for the latitude region above 40N, large, positive, and significant estimates of a change in trend are obtained. Summary results of the 40-60N latitude regions for the analysis, without including the AO index give average pre-turnaround trend (until 1996) of about -1.0 DU/year, change in trend (assumed to start in 1996) estimate of +1.8 DU/year, solar flux effect of 2.3 DU per 100 flux units, and QBO effect of 1.2 DU per 10 knots wind. When AO index is included in the model, it shows a very strong influence on total ozone for latitudes above 35N. Corresponding summary results of the 40-60N latitude region for this case give average preturnaround trend of about -0.89 DU/year (about 0.1 less negative than without AO), change in trend of +1.2 DU/year (about 0.6 less positive than without AO), solar flux effect of 3.8 DU per 100 flux units (about 1.5 more positive than without AO), and QBO effect of 1.1 DU per 10 knots wind (about 0.1 less than without AO). Thus, the change in trend estimates, as well as the solar effect estimates, are impacted by inclusion of AO index, whereas preturnaround trend and QBO effect estimates are impacted very little. Nonetheless, in either case the change in trend estimate shows a substantial positive result, which might provide some mild/limited initial evidence in support of possible beginning of turnaround in total ozone trend.
Analysis of Satellite Total Ozone Data for Detection of Trend Turnaround and Recovery. Applying similar models for ground data to the satellite total ozone, estimation results of the various trend and exogenous predictor variable effects on ozone can now be compared between the ground station and satellite data sources, as well as between the two distinct satellite data sets. In particular, we find that estimation results for the merged TOMS/SBUV are qualitatively similar to results for the ground data. In particular, omitting the data for the 6-month period (12/92-5/93) most influenced by volcanic aerosols from Mt. Pinatubo, the model including the AO index gives results for the 40-60 latitude region a preturnaround trend estimate of .93 DU/year, change in trend of 2.2 DU/year, solar effect of 1.2 DU per 100 flux unit, and QBO effect of 1.0 DU per 10 knots wind. A technical report combining detailed results for the analysis of ground station total ozone data and the analysis of satellite total ozone data is in preparation.
Trend Analysis of Upper Stratospheric Ozone Data for Evidence of Turnaround. In this analysis, both cumulative sum of residuals and change-in-trend model analyses were performed. The data over the three Umkehr stations show a relatively consistent feature of mildly positive tendency in values for the most recent several years. Trend analyses of these data yielded significant positive change-in-trend estimates of about 0.75 percent per year since 1996, relative to negative trends of about -0.55 percent per year existing prior to 1996. This leads to a (nonsignificant) estimate of about 0.2 percent per year positive trend for the recent 5-to-6 year period since the proposed 1996 change-in-trend date. Hence, the analysis provides some preliminary evidence for possible slowdown in negative ozone trend and even positive turnaround in ozone trend over recent years around the 40 km altitude region. The principal investigator for this work was Gregory Reinsel; some details of this work are discussed in Reinsel (2002).
Statistical Diagnostic Analysis of Numerical Models. We recently have applied this two-stage iterative approach to validate the UIUC 2D model for ozone trend analysis with latitudinal zonal averages of TOMS total ozone data from 1978 to 1993. Using ozone output from an earlier version of the model, the results effectively have demonstrated that, whereas the impact of solar cycles on ozone adequately has been captured by the model, there were clear indications of discrepancies in trend changes and seasonal variations. These discrepancies largely have been eliminated after two iteration cycles. Details of this work are given in a technical report currently in preparation.
Combining Information From Statistical Analysis and Numerical Models. The UIUC 2D numerical modeling of the chemistry and physics of the global atmosphere has been contributing in a number of different ways to this topic. Completed and ongoing studies are described below:
• Results from the 2D model have been used as a "model" of expected behavior in statistical analyses of past trends in ozone, based on satellite and ground-based data sets. A paper now is being modified for submission to Geophysical Research Letters.
•Although the process of making improvements to the 2D model continues, results from the current version of the model have been employed for updating the past statistical trends study. These results also are being used in a study to begin looking at the recovery of ozone. A copy of the current 2D model has been provided to the collaborating statistician so that he can perform further tests with the model in support of these studies.
•We are examining how the statistical analyses of observational ozone data can be used to guide further improvements in the 2D model. The analyses of observations are defining key uncertainties in the atmospheric model that are now being investigated.
•Work has begun to conduct an analysis of Equivalent Stratospheric Chlorine Loading as an additional proxy for the expected effects of chlorofluorocarbons and other halocarbons on stratospheric ozone. These results will be included in a future paper on ozone trends and the ozone recovery by principal investigator Don Wuebbles, and postdoctoral research associate, Serge Guillas.
Future Activities:For the next reporting period, we plan to concentrate our efforts on the following two main topics:
Detection of Turnaround and Recovery in Atmospheric Ozone. The focus of this topic is continued investigation of atmospheric ozone data for detection of turnaround in trend and initial signs of recovery in ozone. We will apply currently existing statistical models and methods for analysis of time series and spatial data, and develop new methods of analysis necessary to address specific scientific and statistical issues related to detection of turnaround in ozone. The research will involve analysis of available databases for total column ozone from Dobson and Brewer ground station networks, and from TOMS and SBUV/2 satellite sources, using data as available, at least through the end of 2003. Stratospheric ozone profile data from balloonsondes and Umkehr stations also will be examined. Rawinsonde and other data sources for stratospheric temperatures also will be examined. Results of this research will include:
• Realistic assessments of the strength of current evidence in support of turnaround in total column ozone and in ozone at specific stratospheric altitude layers.
• Assessments of the extent of additional data (length of time) into the future required for detection of an ozone turnaround or start of recovery with specified confidence.
• Comparisons among different ozone measurement systems to assess their capabilities for trend detection.
• Assessment of the impact of various additional explanatory variables on trend behavior of ozone and on trend detection.
• Evaluation of the role of the separate gases affecting ozone trends, including the role of non-halocarbon gases (such greenhouse gases as methane and nitrous oxide) in determining the recovery rate of ozone. Atmospheric modeling studies described below will contribute to these studies.
• Analysis of temperature trends in the stratosphere, and evaluation of the role of temperature trends in evaluating the recovery of ozone. Atmospheric modeling studies also will contribute to these studies.
Atmospheric Modeling and Statistical Analyses. We plan on making an extensive and more thorough analysis of the advantages of combining data analyses with numerical atmospheric model results. It is expected that these analyses will have a major impact on the study of ozone trends in future international assessments of ozone, such as those sponsored by the World Meteorological Organization (e.g., WMO, 1999, 2003). Because these assessments provide the scientific motivation and background for both existing and future considerations of policies to protect the ozone layer (such as the Montreal Protocol, the appropriate portions of the U.S. Clean Air Act, and other policy considerations at the U.S. Environmental Protection Agency [EPA]), this work will be highly relevant to the interests of the EPA. We will focus on the following specific questions:
• What enhancement to understanding past trends in ozone, both globally and regionally, can be derived by using atmospheric model results in the analyses? Will uncertainties in evaluating trends be reduced through the use of atmospheric model results in the analyses of trends? Can the atmospheric model results provide information to separate the underlying causes (dynamical versus chemical) of the observed trends?
• What can the combination of statistical analyses and model results tell us about the expected rate of recovery of stratospheric ozone on both a global and more regional basis? Can turnaround (ozone decrease halted, followed by ozone increase) more readily be evaluated through the combination of data and atmospheric model results?
• What can the statistical analyses tell us about the capabilities and limitations of the atmospheric models? Can the defined limitations help lead to further enhancements of the treatments of chemistry and physics in the atmospheric models? In other words, can the statistical analyses provide direction for further enhancing our understanding of the underlying changes in processes causing the trends in ozone?
• Existing studies have focused on globally averaged datasets, and soon we will expand to evaluating the trends as a function of latitude band. Our existing studies are using the UIUC 2D model. As the studies proceed, we expect to use results from the three-dimensional (3D) models that are being used at the University of Illinois, either the MOZART-3 model being developed in coordination with the National Center for Atmospheric Research or the National Aeronautics and Space Administration Global Modeling Initiative model.
As we move more into the area of 3D model comparison, we will seek to develop innovative statistical tools that relate the global pattern of observed ozone change from the National Weather Service daily global total ozone analyses since 1979 to that of the models. The determination of where the models and data agree or disagree will offer insight into the actual ability of the models to capture the chemical/dynamical system.
Journal Articles:No journal articles submitted with this report: View all 26 publications for this subproject
Supplemental Keywords:atmosphere, ozone, water, watersheds, stratospheric ozone, chemical transport, ecological effects, particulates, environmental chemistry, environmental policy, Great Lakes, EPA Region 5, air quality, health effects, regulation, ecosystem sustainability, air, economic, social, behavioral science research program, ecosystem protection, environmental exposure, risk, geographic area, human health, applied math, statistics, chemical mixtures economics, decision making. , Ecosystem Protection/Environmental Exposure & Risk, Economic, Social, & Behavioral Science Research Program, Air, Geographic Area, Scientific Discipline, Health, RFA, PHYSICAL ASPECTS, Ecosystem/Assessment/Indicators, Engineering, Chemistry, & Physics, Risk Assessments, decision-making, Environmental Statistics, Great Lakes, Applied Math & Statistics, Health Risk Assessment, Physical Processes, Ecological Risk Assessment, Environmental Engineering, Ecological Effects - Human Health, EPA Region, particulate matter, Ecological Effects - Environmental Exposure & Risk, Ecosystem Protection, Monitoring/Modeling, Environmental Monitoring, risk assessment, trend monitoring, ozone , chemical transport models, particulate, risk management, stochastic models, statistical methodology, air quality, computer models, ecological risk, ecosystem health, environmental indicators, ozone, chemical transport, health risk analysis, human health risk, monitoring, policy making, statistical models, particulates, regulations, statistical methods, watersheds, Region 5, air pollution, stratospheric ozone, data models, exposure, water, chemical transport modeling, ecological models, ecological effects, ecological health, human exposure
Relevant Websites:
http://galton.uchicago.edu/~cises/ 
Progress and Final Reports:
Original Abstract
2004 Progress Report
2006 Progress Report
Final Report
Main Center Abstract and Reports:
R829402 Center for Integrating Statistical and Environmental Science
Subprojects under this Center:
(EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).
R829402C001 Detection of a Recovery in Stratospheric and Total Ozone
R829402C002 Integrating Numerical Models and Monitoring Data
R829402C003 Air Quality and Reported Asthma Incidence in Illinois
R829402C004 Quasi-Experimental Evidence on How Airborne Particulates Affect Human Health
R829402C005 Model Choice Stochasticity, and Ecological Complexity
R829402C006 Statistical Approaches to Detection and Downscaling of Climate Variability and Change