From Wednesday 31 July, you’ll see a new format for the Mountain forecasts that MetService produces under its commercial contract with the Department of Conservation.
While the current mountain forecasts for various National Parks are short and to the point, they rely to an extent on the user being able to “join the dots”. One of the main aims of the new format is to be as clear as possible about those weather conditions influencing safety in the outdoors.
Each mountain forecast will contain:
a general statement about the next five days of expected weather – including snow and severe weather
for a selected group of locations in each Park, in addition to the general statement there will also be three days of detailed information on:
as well as a further two days of expected weather for planning
Here’s a guide to what the new format will look like when it goes live on our website:
Can’t read the fine print in the screenshot above? Here’s what it says:
At altitudes between sea level and below the free air freezing level, precipitation falls as rain.
At altitudes above the free air freezing level, precipitation falls as snow and quantity is approximately ten times the water equivalent. That is, 1 mm rain corresponds to about 1 cm snow.
However, the distance to which snow falls below the free air freezing level can vary between a few tens of metres and about 1000 metres, depending upon certain weather conditions. This snow level will be indicated in the worded part of the forecast.
Indication of severe gale may be found in the spot forecasts and/or the worded forecasts.
The layout of the forecasts on MetService’s mobile site m.metservice.com is slightly different (see below) because of the screen size of a smart phone. But the forecast will contain the same information as on www.metservice.com.
The spot locations within each Park for which detailed forecasts are available have been selected in consultation with DOC. Here’s a map of them:
There are a couple of notable changes from the current set of mountain forecasts:
Tararua Forest Park will have its own forecast.
Te Urewera National Park (for which forecasts are currently produced only on a Friday) will now be covered daily by the Brief New Zealand Mountain Forecast (see below).
As is currently the case, forecasts are issued routinely at or before 07:30am and 11:50am and may be updated at any time.
The starting point for an accurate forecast is an accurate representation of the atmosphere’s current state. This is why MetService operates extensive observing and weather modelling programmes and runs a large 24/7 forecasting operation. For much more about this, see the blog on MetService’s Investment in Forecasting and TVNZ’s recent Breakfast programme: “How does the weather work”.
The 7-day rain forecast on www.metservice.com is very useful for planning purposes, with the maps showing pressure, direction and strength of wind, and where rain is expected to fall. The forecast is in the form of a series of maps, which are also provided in a player, so you can animate the series of maps to see how the situation could change over the period of time.
Our 3-day model provides a closer look at New Zealand. On these charts, the thick purple line is a guide to where precipitation may fall as snow.
However, computers deal only in numbers and don’t give the complete story – this is where human expertise comes into play. Trained forecasters do things that weather models cannot, for example:
Consider, in real time, conflicting information – from models and observations – and determine which outcomes are more likely, based on an understanding of the weather situation
Steer a steady course when the model forecasts are “jumping” from run to run
Reconcile observed and forecast weather and recognise when the forecast needs changing (regardless of what the models say)
Explain the weather, particularly to those who are managing weather-related risks (people talk to people; computers talk to computers)
It is vitally important for forecasts (and updates as they happen) to be communicated in a measured, timely and credible manner.
On Wednesday 10 August, days before the outbreak, MetService published a media release and contacted farming organisations, to alert those with a real need to plan in advance about what forecasters were anticipating. Snow and cold would have a large impact to farmers in the lambing and calving season so this was highlighted in the news release.
Because of the chaotic nature of the atmosphere (and the variation of predictability with time), the forecast for a given day (or time) in the future can change as we get closer to it. This is why the forecast issued on a Monday for the weekend to come might be very different to the forecast issued on, say, the Thursday of the same week.
This polar blast was long lived and delivered snow to many regions of the country. While heavier snowfalls have been recorded before in some regions, this is undoubtedly the most widespread and prolonged event since 1939 – the subject of a blog by Erick Brenstrum.
It’s early days in a weather event which is likely to be memorable for its coldness.
Below is a satellite image for midday Sunday. The wind flow over New Zealand is generally from the southwest; the coldest showery air has made its way onto Fiordland, Southland, Otago and south Canterbury.
It’s not raining or snowing over all of southern New Zealand because the wind flow is more or less lined up with the South Island, thereby sheltering some places, and because the precipitation is showery.
The surface temperature doesn’t necessarily tell the whole story. Below is a graph of the air temperature for the 24 hours from 1:00pm Saturday 13 August to 1:00pm Sunday 14 August. Note how the air temperature at Invercargill Airport and Nugget Point, both fairly open to the southwest, has been on a pretty steady downward trend. At 2:00pm Sunday, the temperature at Nugget Point was 1.0 C and the wind speed was 85 km/h. Brrr.
At Timaru Airport, on the other hand, the temperature rose sharply when the southwest change arrived mid Sunday morning – the reverse of what might be expected. The period overnight Saturday to dawn Sunday at Timaru Airport was one of clear skies and fairly light winds, so the quickly-cooling land surface during this period cooled the air immediately above it and an inversion formed. When the southwest change arrived, the air once again became well-mixed. But this southwest air is expected to also become steadily colder.
4:15pm Sunday 14 August 2011
Here’s another look at why the surface temperature doesn’t necessarily tell the whole story.
In my blog post about the winter storm of early July 2011, I partially explained how showers may form in cold air moving over a warmer sea surface. The (relatively) warm sea heats “blobs” of the air immediately above it; these blobs then ascend because they are less dense than surrounding air. For the ascending process to continue, the surrounding air must remain relatively cooler than the ascending warm blobs. Thus, it is important to have information – that is, observations and forecasts – about the vertical temperature structure of the atmosphere.
Observations of the temperature structure of the atmosphere are primarily made using weather balloons. Below is a graph of the temperature at three levels in the atmosphere above Invercargill, obtained from radiosonde balloon flights. The blue line is the temperature at about 5000ft, the red line is the temperature at about 10000ft, and the green line is the temperature at about 18000ft. These heights are approximate; the height of a given pressure level varies with the air temperature; here, we should probably discuss the idea of the thickness of an atmospheric layer – but I think we’ll do that some other time.
Anyway, the graph shows that the atmosphere above Invercargill has been cooling off steadily since the middle of Saturday 13 August. In depth, it is now very cold.
6:15pm Sunday 14 August 2011
It snowed quite heavily in Wellington City, above about 100 metres, from approximately 4:30pm for at least an hour. This is the heaviest and most widespread snowfall in Wellington City for at least 30 50 years.
At midday Sunday 14 August the freezing level around Wellington, obtained from the Paraparaumu radiosonde balloon flight, was just over 1000 metres and falling (it was around 1600 metres at midnight Saturday 13 August). But late on Sunday afternoon, it would still have been well above the level to which snow fell in Wellington. Snow starts melting once it falls below the freezing level – but the melting process draws heat from the surrounding air, which lowers its temperature; thus, the melting snow “drags” the freezing level down with it, at least for a while. How far the freezing level within the area of falling snow is dragged towards the ground depends mostly on the intensity of the snowfall and the vertical variation of temperature and humidity of the air it is falling into.
1:00pm Monday 15 August 2011
So far in this blog, I’ve been talking quite a bit about the temperature throughout the depth of the troposphere (the troposphere is the part of the atmosphere in which weather systems exist). Time, now for a picture. Below is a plot of:
Forecast temperature (colours) at the 500 hPa level (roughly 18,000 ft, or about halfway up the troposphere)
Forecast wind speed (black lines) at the 250 hPa level (near the top of the troposphere)
… for midday Monday 15 August.
The colours in this plot are forecast temperature; over most of New Zealand, the temperature at around 18,000 ft was forecast to be -30 C or lower. The important thing to note is that a large mass of Antarctic air covers almost all of New Zealand.
The red arrow on this plot shows the forecast position of the axis of strongest winds, near the top of the troposphere, at midday Monday 15 August. This is the polar jet, on the border between the deep pool of Antarctic air over New Zealand and the warmer mid-latitude air around it.
Incidentally, the forecast temperatures compare very well with the observed temperatures at midday Monday 15 August, as shown in the table below.
Forecast 500 hPa temperature (C)
Observed 500 hPa temperature (C)
4:30pm Monday 15 August 2011
Here’s a graph of how the freezing level over New Zealand has changed over the last few days.
As of midday Monday 15 August, the freezing level varied between about 1000 ft at Invercargill to about 2000 ft at Whenuapai. Snow has fallen to sea level in many parts of southern and central New Zealand – that is, to at least 1000 ft below the freezing level.
This is a classic example of the melting effect (see the post made at 6:15pm Sunday 14 August 2011, above). Over the last few days, MetService’s Severe Weather Forecasters have spent a lot of time considering how far below the freezing level snow would fall. This requires a good understanding of cloud physics.
9:30am Tuesday 16 August 2011
Here’s a few photos from the Wellington snow of June 1976.
10:30am Tuesday 16 August 2011
In southerly flows, the West Coast of the South Island is well sheltered by the Southern Alps. Since the southerly took hold on Sunday, the air on the West Coast has been very dry because of the Foehn Effect.
Below is a graph of the dew point temperature (the temperature which air must be cooled to for water vapour to condense into water liquid or water solid) at Hokitika Airport from 10am Sunday 14 August to 10am Tuesday 16 August. On the afternoon and evening of Sunday 14 August, there’s a huge change in dew point (around 13 degrees), down to around -10 C. Since then, the dew point has remained negative, generally fluctuating between about -3 C in the morning and -7 C in the afternoon. Such a low dew point makes the air feel much colder than its temperature would suggest. We take the dew point into account when calculating the “feels like” temperature.
2:30pm Tuesday 16 August 2011
This event has been characterised by many places having low daytime (maximum) temperatures.
Maximum temperature on
Lowest daily maximum temperature on record
Month / year occurred in
Monday’s max temperature is the lowest since …
15 August 2011
New Plymouth Airport
25 July 2011 and 12 July 1951
Napier (Nelson Park)
25 July 2011 and 17 July 1995
Monday night / Tuesday morning was very cold in some places, though. Here’s a few notable overnight minima from MetService automatic weather stations.
Waiouru Automatic Weather Station
-7.7 C (new record for August)
Blenheim Airport Automatic Weather Station
-6.2 C (new record)
Rotorua Airport Automatic Weather Station
-5.2 C (equals record)
Taupo Airport Automatic Weather Station
5:00pm Tuesday 16 August 2011
Below is a plot of where the air arriving at an altitude of 500 metres above Auckland at midday Monday 15 August came from. Four days previously, it was over the Antarctic landmass; two days previously, it was still over the Antarctic sea ice. The Antarctic sea ice edge is close to its northern-most extent and is near latitude 60 degrees South. Thus, the air arriving at Auckland passed very quickly over the relatively warm ocean between the Antarctic ice edge and New Zealand. In contrast, the air from the Southern Ocean which arrived over Auckland on Saturday 9 July (see my blog post on the stormy period of early July 2011) had travelled over a much longer stretch of ocean, over a longer period of time, and consequently was warmer and moister.
6:00pm Tuesday 16 August 2011
As of 2:00pm Tuesday 16 August 2011, the extent of snowfall in this storm is as shown in the image below.
4:00pm Wednesday 17 August 2011
During the next few days, while an anticyclone advances onto the country, the general wind flow will decrease in strength and the depth of cloud along eastern coasts gradually reduce. Near sea level the air over New Zealand remains very cold, and the advancing anticyclone more or less “traps” it in place. Very cold air, clear skies and light winds overnight are a recipe for hard frosts.
Hopefully, the diagram below – known technically as a tephigram – helps illustrate this. It is a plot, in the vertical, of the air temperature and the dew point temperature derived from the radiosonde balloon flight at Invercargill at midday Wednesday 17 August. At Invercargill there is already a large mass of sinking, warming (and drying) air above about 5000 ft (see text in red on diagram). This sinking air presses on the (relatively) colder air beneath it, trapping it near the Earth’s surface. In this particular case, the zone of transition between the two different air masses is known as a subsidence inversion. I’ve marked the subsidence inversion on the diagram; it’s the broad blue horizontal bar near the bottom.
The very cold air trapped below about 5000 ft at Invercargill is much less inclined to move around than the air further up in the atmosphere. As I’ve explained above, this is partly because of the advancing anticyclone. But it’s also partly because cold air is less “runny” than warm air. (Treacle flows much more readily when warm than cold). On the right of the diagram below are the winds in the vertical, as they were above Invercargill at midday Wednesday 17 August. Clearly (see text in green on diagram), the one wind barb shown below 5000 ft indicates quite a different flow from all the winds above 5000 ft: the flow near the surface has become decoupled from that above.
In a general sense, this vertical temperature and wind structure is expected to spread over eastern parts of New Zealand during the next few days as the anticyclone moves closer and pressures over the country rise.
Forecast surface pressure field for midnight Wednesday 17 August 2011.
Forecast surface pressure field for midnight Thursday 18 August 2011.
3:30pm Friday 19 August 2011
Finally today, cloud over the south of the South Island has cleared enough to reveal the extent of snow cover there.
Below are two visible satellite images. The first is for around 10:00am on the morning of Wednesday 10 August, some days before this extraordinary cold outbreak. The second is for around 10:00am on the morning of Friday 19 August. Nearly all of the white over Canterbury, Otago, Southland and Fiordland is snow. The imagery only shows the extent of the snow, not its depth.
In New Zealand, the coldest days of the year are usually in late July or early August.
The most significant cold snap since 1995 gripped the nation from late Friday 22 July until late on Monday 25 July. Significant amounts of snow fell in the south and east of both islands, with over 30 cm to near sea level reported in the Christchurch area. Many of the arterial routes through the North and South Islands were closed, including the Waioeka Gorge route from Gisborne to Opotiki.
This cold episode was the result of a river of air flowing straight from the Antarctic ice shelf to New Zealand’s shores. For this to occur, a high pressure zone over Tasmania needs to occur at the same time as a low pressure system deepens over the Chathams area, so that between them a southerly flow extends from the Antarctic to New Zealand for a long enough time to allow polar air to reach the country.
As the weather map above shows, the “polar blast” was just arriving over the far south of the country at midnight Saturday 23 July (behind the old front); it spread across the country during the following two days. Because it is so cold, polar air is generally very dry. But as it passes over the (increasingly warmer) ocean between the Antarctic and New Zealand, it takes up heat and moisture from the sea surface: this is what drives the formation of showers.
Had the polar air arrived a day or so earlier it would have encountered warmer air associated with the low which brought heavy rain to parts of Northland, Bay of Plenty and Gisborne a few days earlier; the result would have been even more snow.
The coldest air made its way across the North Island early on Monday 25 July and a trough following on behind brought a further bout of heavy snow later that day. Heavy showers brought the snow level down as they passed: a centimetre or so of snow even fell in Greytown (about 100 metres above sea level) in the Wairarapa to the east of Wellington.
This blog post is the third in a three-part series discussing verification of MetService forecasts. Here, we present the method used for verifying Severe Weather Warnings, along with some recent examples.
The statistical method of verifying Severe Weather Warnings is similar to that used for verifying rainfall in city forecasts. That is, a categorical approach is taken, with Probability of Detection (POD), False Alarm Ratio (FAR) and Critical Success Index (CSI) being produced on a monthly basis. In the case of Severe Weather Warnings, POD, FAR and CSI are expressed as percentages.
There is, however, much more subjectivity in determining whether a Severe Weather Warning is a success, a false alarm, or whether an event has been missed. This is because Severe Weather Warnings apply to broad areas (minimum of 1000 square kilometres, and generally much larger than that) and an assessment has to be made of how widespread the occurrence (or non-occurrence) of heavy rainfall / heavy snowfall / strong winds was. The initial assessment is made by MetService’s Severe Weather Forecasters themselves, using all available weather observations (including those from many voluntary observers MetService can call), media reports, comments on web forums about New Zealand’s weather, radar rainfall accumulations, and occasionally through discussion with territorial authorities and utility companies. At the end of each month, every assessment of every event is reviewed by two senior meteorologists who work at MetService and who are outside of forecasting operations, and I sign off on them as Chief Forecaster.
Rainfall, snowfall and wind are usually highly variable over any broad area in New Zealand – because of the complexity of the New Zealand landscape and because of the variability of weather systems themselves. For example, if the Severe Weather Warning is for 150mm rain in 24 hours, is it successful if:
The rain fell in 30 hours? 18 hours?
200mm fell? 100mm? …
250mm fell over half the area and 50mm fell over the other half of the area?
Performance over the year to September 2010
Graphs like those above but covering a much longer period – say, a decade – show a gradual improvement in forecast accuracy. Given the changes in observing technologies, numerical weather prediction and forecasting techniques over the last several decades, this is not surprising. What these verifications don’t show well is the much increased precision – location, amount, intensity, timing – that Severe Weather Warnings have contained in recent years.
MetService’s performance targets for Severe Weather Warnings are:
Probability of Detection
False Alarm Ratio
It is possible to have a “perfect” POD or FAR – but not simultaneously:
If heavy rain (or heavy snow, or strong winds) was forecast every day, the POD would be 100% – but the FAR would be close to 100% too.
If heavy rain (or heavy snow, or strong winds) was never forecast, the FAR would be 0% – but the POD would be close to 0% too.
In both of these cases, the forecast would be of no value because users would have no idea on which days heavy rain (or heavy snow, or strong winds) was expected to occur.
Very few forecasts or warnings are issued in isolation. Therefore, verifying them in isolation does not provide a complete picture of their value. Not infrequently, MetService will issue a Severe Weather Watch if it is considered that an event will not quite meet the Severe Weather Warning criteria but is nevertheless notable. If, subsequently, the Severe Weather Warning criteria are met, the verification scheme will record a missed event – despite MetService having perhaps come very close to forecasting it perfectly.
Today MetService has issued a severe weather warning for heavy snow over parts of South Island. Looking back over the past week, it is unusual for a single weather event to result in severe weather warnings for all of widespread heavy rain, severe gales and heavy snow. But this event has been far from typical in its severity and longevity.
The media have extensively reported the full impacts of the storm, so I won’t go into that here. Instead I’ll look a little more deeply into the meteorology of what’s been going on, starting from day one.
Here is an animation of the surface weather maps starting midnight on Friday 21 May – frames are every six hours throughout the event, and the isobars are at 2 hPa spacing:
There is a lot going on in these surface maps. To give a full account of what’s been happening I should really include a discussion about the weather in the upper air – but to keep this post short I’ll concentrate just on the surface here. Points to note from the maps are:
at the start of the loop, an old Low over the south Tasman Sea weakens
a new Low forms off the south Queensland coast, then moves southeastwards towards New Zealand and deepens
a second Low centre forms west of North Island, then a third forms over Bay of Plenty
the third Low takes a very unusual track, moving south then southwestwards towards Canterbury
a High moves onto the south Tasman Sea and stays there for the rest of the period
a very strong east to southeasterly airstream develops over the bottom half of South Island
the Queensland Low eventually moves eastwards over North Island
The centre of the Queensland Low took a rather convoluted track as it approached us:
Track of the low that formed on 21 May 2010
I guess you could say it “looped the loop” near longitude 170E. This was caused by the formation of the secondary and tertiary Low centres that shifted the “centre of gravity” of the broader system away from the originating Low.
If you look again closely at the loop of weather maps above, you’ll notice that just after halfway through, the second and third Lows are dumb-belling cyclonically around each other over New Zealand. This motion occurs when a multi-centred Low develops – we sometimes call these systems complex lows. The combined motion plus the spiralling bands of precipitation were well captured by our current radar network.
The following animation shows where our weather radar is detecting rainfall-sized drops, and the shape of the rain areas illustrates the position and movement of the Lows. Pictures are seven minutes apart and cover the period from the evening of 25 May to the early hours of 26 May. Light falls are yellow and heavier falls blue.
Click to view animation. Note – the animation is a large gif file: 2.9MB
The radar pictures also show the change in texture from north to south, with the northern precipitation looking more speckly (showery) and the southern precipitation more uniform (rainy).
As I mentioned at the beginning, this storm is not over yet, with snow still to come in the south. You can keep up to date with the latest weather by continuing to check out metservice.com.