Air quality update 4. Week two of level 4 restrictions

Changes in NZ air quality due to COVID-19 level 4 lockdown

Update #4 – second full week of level 4 restrictions

Tuesday 14th April 2020

Headlines

• The large reductions in traffic pollution in our major cities seen during week 1 of lockdown have been maintained during week 2.
• Reductions in traffic pollution of at least 70 % at most sites broadly match reductions in traffic volumes released by the NZTA.
• Levels of PM₁₀ (particulate matter small enough to be inhaled deep into the lungs) are down by 37 % on average across Auckland relative to the norm for this time of year. This is greater than the normal fluctuation of levels in PM₁₀ of +/- 20 – 30 %.
• Whereas we would normally expect PM₁₀ to start rising at the start of April due to the beginning of the home heating season, there is no clear evidence of this in Auckland yet, possibly due to the warm weather during the lockdown period so far.
• Across the towns of lower South Island (from Timaru to Invercargill) PM₁₀ levels usually start rising in April due to smoke from the burning of wood or coal for home heating. PM₁₀ here is currently approximately normal for the time of year, but we’ll be keeping a watch on any abnormal increases due to more people being at home.

Why review after a second week?

Air quality can fluctuate greatly between one day and another due to changes in the weather, and especially whether a monitoring station is upwind or downwind of nearby major roads. This means simply comparing one day’s air quality with yesterday’s can be misleading.

We published an initial analysis of the traffic exhaust pollutant – nitrogen oxides - at the end of week 1, however the weather was probably not typical during that week. With an additional week of data, we have now experienced a wider range of weather conditions and the results are therefore a more reliable indicator of changes in emissions as a result of the lockdown.

We’re comparing air quality now with ‘normal’, but how do we define ‘normal’?

As well as fluctuating daily, air quality changes on a seasonal basis, being generally worse in winter and better in summer. However, for most pollutants, the long-term trend is very slow so that concentrations have changed relatively little from year to year over the past 5 years. To account for these effects, this analysis compares hourly data from the first and second week of level-4 lockdown with average air quality in the same weeks (14 and 15), and one week either side, for the previous 5 years, which we take to represent ‘normal’ air quality at each site (unless there is evidence of a significant change in the ‘normal’ in the 5 years preceding the lockdown).

Since the week 1 update, the method has been slightly refined and the week 1 figures have been recalculated leading to very small differences compared to what we previously reported.

Data sources:

Routine automated air quality monitoring is conducted by regional councils. This analysis is based on unverified data at the moment and conducted by NIWA Principal Air Quality Scientist Dr Ian Longley. Traffic data was taken from https://www.nzta.govt.nz/about-us/coronavirus-disease-covid-19-services-update/weekly-traffic-count-information/.

What this analysis does, and does not, focus on

This analysis updates our previous report on concentrations of oxides of nitrogen (NO_x). NO_x consists of two pollutants (nitric oxide and nitrogen dioxide). They are known to exacerbate asthma, stunt childhood lunge development and are linked to numerous other health impacts for the young and vulnerable. In our cities road traffic exhaust is overwhelmingly the main source, especially diesel vehicles, and so NO_x can also be used to indicate levels of all traffic exhaust pollutants.

This analysis is also our first to look at levels of particulate matter (PM).  As well as traffic exhaust, PM also comes from road dust, natural sources like sea salt, and (during winter) smoke from the burning of wood and coal especially for home heating. PM₁₀ (particles smaller than 10 microns and likely to be inhaled deep into the lungs) is monitored at more than 50 sites across the country. Levels in some towns, mostly in South Island, can breech National Environmental Standards on cold winter days.

What are our findings?

1. Nitrogen oxides (traffic exhaust pollutants)

The reductions in NO_x concentrations, averaged over each week, are presented in table 1. There is very little difference between week 1 and 2. This corroborates our week 1 analysis and also shows that the large reductions in traffic pollution in our major cities seen during week 1 of lockdown have been maintained during week 2.

Table 1: Changes in week-average NO_x concentrations compared to the norm for this time of year (based on 2015-2019).

	Week 1	Week 2	Week 1 & 2 combined
Queen Street, AKL	-61 %	-59 %	-60 %
Henderson, AKL	-65 %	-71 %	-68 %
Takapuna, AKL	-68 %	-75 %	-72 %
Penrose, AKL*	-49 %	-72 %	-61 %
Willis Str, WTN	-76 %	-74 %	-75 %
Riccarton Rd, CHC	-79 %	-78 %	-79 %

*not included in our previous update

In our week 1 update we did not report the degree to which weekly NO_x concentrations may normally fluctuate due to weather variations. This is approximately +/- 40 % for NO_x. The fact that reductions higher than this have been reported consistently for two weeks at all sites allows us to conclude that the reductions during the 2 weeks of lockdown are caused mostly by reductions in traffic, not changes in the weather.

The reductions are of a similar magnitude to reductions in traffic during week 1 as reported by NZTA on 8^th April (table 2). These reductions relate to particular locations on the State Highway network which do not match the location of air monitoring stations, but may be taken as indicative of each city in the absence of more detailed information.

Table 2: Average week 1 reductions in NO_x and traffic data provided by NZTA

	Average NOx reduction	Average traffic volume reduction
Auckland	-65 %	-76 %
Wellington	-76 %	-82 %
Christchurch	-79 %	-77 %

It is still too early to fairly compare these reductions to those observed overseas as not all data is available or comparable. However, one recent analysis has indicated reductions in the UK over the period 1^st – 30th March of 20 – 40 %. The larger reductions observed in Auckland may be due to differences in the extent of lockdown and traffic reduction, but also because of New Zealand’s isolation which means that we are not exposed to air pollution from neighbouring countries.

2. Particulate matter (traffic, heating, industrial and natural sources) in Auckland

The reductions in PM₁₀ concentrations, averaged over each week, are presented in table 3. Like NO_x, there is very little difference between week 1 and 2. Reductions are smaller than for NO_x but this is almost certainly because a large proportion of PM₁₀ in Auckland at this time of year is sea salt generated by breaking surf which will not be impacted by the lockdown.

Table 3: Changes in week-average PM₁₀ concentrations compared to the norm for this time of year (based on 2015-2019).

	Week 1	Week 2	Week 1 & 2 combined
Queen Street, AKL	-32 %	-44 %	-38 %
Henderson, AKL	-31 %	-38 %	-35 %
Takapuna, AKL	-45 %	-42 %	-44 %
Penrose, AKL	-26 %	-39 %	-33 %
Glen Eden, AKL	-26 %	-30 %	-28 %
Papatoetoe, AKL	-36 %	-37 %	-37 %
Patumahoe, AKL	-18 %	+1 %	-9 %

*based on one year of historic data only (2018)

Concentrations of PM₁₀ in urban Auckland usually fluctuate above and below season-average concentrations by 20 – 30 %, largely due to fluctuations in the weather. During week 1 and 2 of level-4 lockdown PM₁₀ concentrations have fallen by 26 – 53 % (average 37 %) as observed at 7 urban monitoring stations across the region. Changes in PM₁₀ at Patumahoe (a reference rural station outside Pukekohe) are well within the normal variation further indicating that the reductions across urban Auckland can probably be attributed to reductions in traffic.

3. Particulate matter from home heating

Whereas we would normally expect PM₁₀ to start rising at the start of April due to the beginning of the home heating season. With more people at home for longer during lockdown, there is a possibility of heating emissions being higher than normal this year.

There is no clear evidence of PM₁₀ rising in Auckland yet, possibly due to the warm weather during the lockdown period so far.

Across the towns of lower South Island (from Timaru to Invercargill) this effect is usually stronger as it gets colder faster. However, PM₁₀ levels are currently approximately normal for the time of year.

We’ll be keeping a watch on this and providing an update later in April.

Read more updates on the changes in NZ air quality due to COVID-19 level 4 lockdown.