Options for research to enhance national air quality information for environmental reporting
Prepared for Ministry for the Environment
June 2015
Prepared by:
Guy Coulson, Elizabeth Somervell, Ian Longley, Sally Gray and Gustavo Olivares
For any information regarding this report please contact:
Guy Coulson
Group manager
Air Quality and Health
+64-9-375 4503
Email:
National Institute of Water & Atmospheric Research Ltd
Private Bag 99940
Viaduct Harbour
Auckland 1010
Phone +64 9 375 2050
NIWA CLIENT REPORT No: AK2015-15
Report date: June 2015
NIWA Project: MFE15102
Quality Assurance StatementReviewed by: / Annette Semadeni-Davies
Approved for release by: / Ken Becker
Contents
Executive summary 5
1 Introduction 6
1.1 Development Process 6
2 Potential improvement projects 7
2.1 Potential Projects: General indicators and data framework questions 8
2.2 Potential projects: indicators of pressure 9
2.3 Potential projects: indicators of state 10
2.4 Potential projects: indicators of impact 11
3 Prioritisation process 13
3.1 Prioritisation criteria 13
3.2 Introducing dependencies into the prioritisation 16
4 Suggested work packages for the improvement projects 17
4.1 Project workflow framework 17
4.2 Process to review and update indicators 18
4.3 Creating indicators of PM2.5 state and impact 18
4.4 Robust estimates of the contribution of natural sources to PM10 22
4.5 Establishment of a national monitoring network 26
4.6 Creation of an easily updated and improved national emissions inventory suitable for tracking trends in pressures 29
4.7 Improving the indicators of PM10 state and impact 34
5 Detailed Work Packages 38
5.1 Creating indicators of NO2 state and impact 38
5.2 What are the meteorological impacts on trends in ambient pollution? 44
6 Completion schedule 50
7 References 51
Tables
Table 31: The project-criteria matrix. 15
Table 32: Projects grouped by dependencies. 16
Table 41: Estimated timings for the PM2.5 work package. 20
Table 42: Estimated timings for the PMn work package. 24
Table 43: Estimated timings for the inventory work package. 31
Table 44: Estimated timings for the PM10 work package. 36
Table 51: Estimated timings for the NO2 work package. 40
Table 52: Estimated timings for the meteorological work package. 46
Table 61: Recommended Sequencing of work. 50
Figures
Figure 41: Anticipated workflow framework. 17
Options for research to enhance national air quality information for environmental reporting
Executive summary
The Ministry for the Environment (MfE) and Statistics New Zealand (Stats NZ) are jointly responsible for the publication of environmental domain reports and the continual improvement of the reporting of environmental measures. The first air domain report was published in 2014. The Parliamentary Commissioner for the Environment (PCE) and the Technical Advisory Group (TAG) set up by MfE as part of the reporting process have both provided suggestions for potential improvements. NIWA was subsequently commissioned by MfE to undertake the development of a prioritised list of projects to enhance and expand the air domain report.
This report presents this list of proposed projects, and documents the process of its development. The process was iterative and involved rounds of consultation with MfE and StatsNZ. An initial list of project ideas based on feedback and expert opinion, was developed by NIWA, who also devised criteria for each potential project to be scored against. The resulting high priority projects were grouped by their interdependencies and re-configured into proposed work packages which align with a common workflow sequence.
The proposed workflow breaks the recommended work package into staged projects, each of which allows for the generation of improved interim outputs (such as case studies, draft indicators) when required. Projects feature a mixture of literature reviews, method development and modelling, and a mixture of campaign and continuous monitoring. Some projects are reliant on original research, or allow for alignment with ongoing research by NIWA or others. Outline costs are provided in those areas where it is possible to do so. In other areas costs will depend on scope which is often defined at various milestones within each work package. We do not propose a detailed implementation plan but indicate that in the case of monitoring especially implementation may be dependent upon other agencies, especially Regional Councils and NZTA.
The list of potential projects can be found in Section 2, the criteria used to score them in Section 3. Section 4 begins with the workflow framework and then describes the work packages developed to complete the high priority projects. These are:
§ Creating indicators of PM2.5 state and impact
§ Robust estimates of the contribution of natural sources to PM10
§ Establishment of a national monitoring network
§ Creation of an easily updated and improved national emissions inventory suitable for tracking trends in pressures
§ Improving the indicators of PM10 state and impact
At the request of MfE and StatsNZ, two work packages have been developed in more depth in Section 5. These are:
§ Creating indicators of NO2 state and impact
§ What are the meteorological impacts on trends in ambient pollution
This report also recommends establishing a National Monitoring Strategy: a working document that outlines the critical monitoring dependencies for the Air Domain Report, to share with stakeholders.
Options for research to enhance national air quality information for environmental reporting 21
1 Introduction
In early 2014 the Ministry for the Environment (MfE) and Statistics New Zealand (Stats NZ) published the 2014 air domain report (MfE and StatsNZ 2014), the first of a series of intended environmental domain reports. This was followed in early 2015 by a report from the Parliamentary Commissioner for the Environment (PCE 2015) commenting on the air domain report, which was part of the process for environmental reporting. The PCE report made several recommendations for improvements to the domain report including clearer objectives and conclusions The domain report’s Technical Advisory Group (TAG) have also commented potential improvements. Continued development to enhance the quality of reporting is an integral part of the environmental reporting framework and accordingly, MfE and Stats NZ have requested that NIWA: consolidate suggestions for improvements; identify any research required; assess the importance of the research; and assess the effort required to carry it out. The objective of this report is to document the development of a prioritised list of potential research projects and implementation plans (work packages) to assist MfE and Stats NZ with their planning for future air domain reports.
1.1 Development Process
The rest of this report is arranged in four sections as follows. Section 2 lists potential projects that could improve the reporting of air quality in New Zealand. The list was drawn up after the release of the PCE recommendations and TAG feedback in consultation with MfE and StatsNZ. Section 3 presents a set of prioritisation criteria that can be used to score the projects. The potential projects were scored as though they were fully completed to take into account the full range of impact each could eventually have. The highest ranking projects were reconfigured into work packages (Sections 4 and 5). These included the essential dependencies required to complete the work. The final work packages described in Section 4 evolved out of a series of meetings and brainstorming sessions. At this point the similarities between the work packages became apparent and a workflow framework (Figure 4-1) was created to illustrate the iterative and flexible nature of most of the proposed work. Further feedback from MfE resulted in extended descriptions of two of these work packages: the NO2 and meteorology work packages. These are described in Section 5. Finally, Section 6 provides a completion schedule for the list of prioritised projects.
2 Potential improvement projects
Consultation with MfE and Stats NZ regarding suggested improvements (informed by improvements identified by the TAG and the PCE) to environmental domain reporting led to the creation of a list of potential projects to improve the air domain report which are detailed in this section.
Many of the suggested improvements are about finer scale reporting, e.g. regional rather than national reporting of emissions or health impacts on Maori, but the data are scarce. Other suggested improvements call for reporting of pollutants other than PM10. Most of the suggestions se do not currently meet MfE and Stats NZ data framework requirements. All such work would need to examine how available data on geographic and demographic factors or multiple pollutants compare to the national data framework. For completeness, all potential improvements are examined in this section, regardless of any on-going work or plans to address these needs.
The exception to this are developments in the harmonisation of reporting. Currently, Regional Councils and other statutory bodies are working on the standardisation of the pollutant measurements they report upon. This will make a valuable contribution to the robustness of national reporting. It is recommended that any work on improvements to environmental reporting keep a watching brief on these developments.
The potential projects have been grouped into four categories. These are listed below and are described in Sections 2.1-2.4:
§ General indicators and data framework questions
− Is an Air Quality Index (AQI) a suitable metric for use as an indicator?
− What are the benefits and shortcomings of single-number indicators to characterise all of New Zealand's air quality?
− Creating and maintaining a national database of air quality measurements and associated information and enabling automated generation of environmental reports. (Currently partly addressed by the development of LAWA)
§ Indicators of pressure
− What are the meteorological impacts on the state and trends in ambient pollution?
− How accurate is a home heating emissions indicator?
− Creation of an easily updated and improved national emissions inventory suitable for tracking trends in emissions
− Primary NO2 emissions
− Improvements to the road vehicle emissions estimates
− How much PM10 is natural?
− Impacts of climate (i.e., indicators in atmosphere and climate domain report) on indicators in air domain report?
§ Indicators of state
− Creating models of unmonitored areas
− Improving understanding of the state of PM2.5
− How does spatial variation of PM10 in Christchurch affect the reported population weighting?
− Improving NO2 coverage
− Creation of a National Monitoring Network
§ Indicators of impact
− New national exposure (and impacts) model scoping study
− Further evaluate the accuracy of HAPiNZ (Health and Air Pollution in New Zealand) and recommend any necessary improvements
− What are the health impacts of NO2, PM2.5, PMcoarse and PMnatural?
− Impacts of air indicators on climate (i.e., indicators in atmosphere and climate domain report)?
2.1 Potential Projects: General indicators and data framework questions
§ Is an Air Quality Index (AQI) a suitable metric for use as an indicator?
An Air Quality Index (AQI) is a composite indicator created from a variety of pollutant metrics. This work will investigate options and suitability of AQIs for reporting air quality in New Zealand. Investigation could include, for example, whether any bias is introduced by particular pollutants due to monitoring bias (peak sites – that is sites where concentrations are known or expected to breach the standards), availability of data etc.
§ What are the benefits and shortcomings of single-number indicators to characterise all of New Zealand's air quality?
Indicators used for the domain report were formulated under unfavourable time constraints without a rigorous and documented decision-making process. The PCE suggests that reporting and hence indicators are considered a “…work in progress…” which will require a review process (e.g. continual or occasional). This work will examine and document the process of choosing and maintaining indicators including the trade-offs between coverage and accuracy. It will use air quality as an example but may be applicable across other domains.
Other suggested improvements call for reporting a range of pollutants other than PM10. Most of these would not currently meet data framework requirements, which was the primary reason that only PM10 was included in the domain report. This work will (briefly) re-assess whether adding more pollutants to the reporting schedule will enhance the information content of the reporting. It will look at which pollutants might be included and which will either not add substantially to the information content or will not meet data framework requirements without major changes.
§ Enabling automated generation of environmental reports
Currently models such as HAPiNZ (Kuschel et al., 2012) are static, any updates to available data have to be individually commissioned and carried out laboriously by hand. This work will investigate the processes required for updates to be carried out on a regular basis and for automating as much of the process as possible.
2.2 Potential projects: indicators of pressure
§ Creation of an easily updated and improved national emissions inventory suitable for tracking trends in emissions
Indicators of pressure all depend to some extent on emissions inventories, but the existing inventories in New Zealand have some limitations in their ability to track trends. Improved national inventories will also enable better understanding of the links between pressure and state at a national scale. This work will create a single national emissions inventory for major pollutant sources, including wood-burning and shipping. It will also assess the requirements for upkeep and maintenance responsibilities.
The need and possibility of regional or other breakdown will also be assessed
§ Improvements to the road vehicle emissions estimates
The Vehicle Emissions Prediction Model (VEPM) is currently un-validated but is the only vehicle emissions model available for the New Zealand vehicle fleet. Some work has been done to evaluate it against on-road emissions measurements using remote sensing (RSD) measurements (Kuschel et al., 2012), and it is possible to extend that validation work. Data-mining and targeted RSD (and other) measurement campaigns may be able to improve (e.g.) diesel versus petrol emissions (Smit and Somervell, 2015). This project would continue and extend initial validation work and rolls into the emissions inventory review described above.
§ Primary NO2 emissions
While emissions of oxides of nitrogen (NOx) are decreasing, measured NO2 concentrations are not. One possible reason for this anomaly is that the proportion of NOx emitted directly from the tailpipe as primary NO2 is increasing. This work will investigate methods to quantify and track trends in direct NO2 emissions from traffic. Results may be used to update traffic emissions models