MetService in the Tropics

Posted on December 10, 2012 by Jonathan Tunster

Beach weather or a tropical cyclone?

If you’re going on holiday in the southwest Pacific, you should check out the MetService TV Pacific Weather update. But it isn’t always “beach weather” in the tropics. Severe weather in the southwest Pacific can have very significant consequences.

Above: Late afternoon cloud build-ups at Henderson Airfield, Guadalcanal, Solomon Islands.
Severe weather on the way?

South Pacific Guidance

Since 1 November 2009, RSMC (Regional Specialized Meteorological Centre) Wellington, of MetService, Lead Meteorologists have produced twice-daily broadscale guidance for expected severe weather over the next 5 days in the tropical southwest Pacific. This type of guidance is known as the South Pacific Guidance charts, or SPG.

Above: From July 2012, an example of South Pacific Guidance produced by a Lead Forecaster at RSMC Wellington (MetService). It highlighted areas and probabilities of severe weather developing 4 days out.

It’s important to note that we are not doing the job of forecasters in each country. This South Pacific Guidance (SPG) chart is another piece of information that they can use to write their forecasts, but should not be used in isolation.

Why do we do this?

SPG charts are one of the key products of the Severe Weather Forecasting and Disaster Risk Reduction Demonstration Project, (SWFDDP). The project is overseen by the World Meteorological Organization (WMO), based in Geneva. What is it trying to do?

Goals for the South Pacific SWFDDP Project

Above: The goals for the project.

The project is not unique to our part of the world. It began in Africa in the mid-2000s, and there are now two separate African projects covering 16 countries in Southern Africa and 6 countries in Eastern Africa. RSMCs in Pretoria, La Reunion, Nairobi and Dar es Salaam lead these projects. There are future plans for similar projects in Southeast Asia and the Bay of Bengal.

The southwest Pacific project includes 9 Pacific Island Countries ( Solomon Islands, Vanuatu, Fiji, Tuvalu, Kiribati, Tonga, Samoa, Niue and the Cook Islands) and 2 Territories (American Samoa and Tokelau) .

What do we mean by Severe Weather?

“Severe weather” needs to be defined carefully and in a way that is worthwhile and meaningful for all participating countries. Differences in latitude and terrain across the SW Pacific make this a tricky task. The current thresholds for severe weather are shown below:

Above: Severe Weather Thresholds for the SWFDDP.

Large waves, strong winds and heavy rain can all exist outside of tropical cyclones, whereas tropical cyclones can include all of the above.
For example, large waves have had a significant impact on low lying atolls in Kiribati and Tuvalu, strong trade winds have contributed to damage in Rarotonga and sinking of a vessel in the Solomon Islands, and heavy rain caused major flooding in Fiji in March 2012; all without the presence of a tropical cyclone.

As shown in the goals above, a major aim of the project is to raise awareness and capabilities in dealing with severe weather outside of tropical cyclones.

The first image in this post showed cloud build-ups about the higher ground of Guadalcanal. These may produce rainfall in excess of 100mm in 24 hours, but if they cover a very small area are not currently included in the SWFDDP. The project is really focussed on broadscale weather, which can be predicted with some accuracy days ahead of time.

What else does MetService do in the tropics?

The Lead Forecaster draws the “SPG” charts we saw above. Other duties of the Lead Forecaster in the tropics include regular discussions with meteorologists in RSMC Nadi regarding future tropical cyclone formation, production of bulletins for the New Zealand Ministry of Foreign Affairs and Trade detailing future tropical cyclone potential, and drawing streamline analyses.

Above: South of 25S, the familiar isobars are drawn. To the north, the teal-coloured lines are streamlines which show the windflow. The double black lines north of the equator mark the ITCZ, a major area of wind convergence. For more on convergence, see http://blog.metservice.com/2012/10/convergence-lines/

Streamline analyses look quite different to the regular “weather maps” shown in New Zealand newspapers, the MetService website, and on TV.

This type of chart is drawn because as you get further into the tropics (especially between the latitudes of 15S and 15N) drawing of isobars becomes increasingly difficult, ultimately impossible, and the wind no longer behaves with the same relation to isobars as it does in temperate regions.

In producing SPG charts, lead forecasters have to be aware of what is happening right now (using satellite data and surface observations and combining these into a streamline analysis) and what is expected to happen during the next 5 days. This requires combining meteorological understanding with numerical weather prediction (NWP) products from supercomputers around the world. In particular, products known as ‘ensemble prediction systems’ are used.

Ensemble products provide a way of dealing with the uncertainty in the atmosphere. The atmosphere is large and extremely complicated, hence it’s impossible to perfectly know everything about it at any given moment. But the problem is that the atmosphere is very susceptible to very minor changes in what is happening right now. These minor changes are proven to have a huge impact in the future. This is known as Chaos Theory and has been popularised with the well-known “butterfly effect”.

Global Ensemble Forecast System forecast tracks for SANDY (courtesy NOAA)

Above: Global Ensemble Forecast System forecast tracks for Hurricane SANDY (courtesy NOAA). Ensemble forecasting aims to find out the impacts of initial uncertainties in the atmosphere on the future conditions of the atmosphere – in this case the position of a hurricane.

Understanding, and interpreting, ensemble prediction systems is a delicate operation, but it is entirely essential in modern forecasting. Without it, forecasters are severely restricted in their outlooks.

MetService also runs a website called MetConnect Pacific. This is a portal for forecasters involved in the SWFDDP to access information, observations and the South Pacific Guidance charts drawn by MetService Lead Forecasters.

Climate Change Development Fund

Thanks to financial aid from the Climate Change Development Fund administered by the New Zealand Ministry for the Environment, MetService has been able to conduct in-country training between May 2012 and November 2012 for all nine participating Pacific Island Countries. On each trip, two meteorologists from MetService have provided training, along with meteorologists from the National Weather Service in Hawaii for the trips to Samoa and Fiji.

What does this project, which focusses on the weather over the next 5 days, have to do with climate change, which looks into seasonal, interannual to decadal changes timescales?

We believe that dealing with extreme weather is the first vital step towards managing the risks of extreme events and disasters to advance climate change adaptation at the community level. According to many climate scientists, climate extremes (extreme weather or climate events) such as substantial warming in temperature extremes, heavy precipitation or the proportion of total rainfall from heavy falls, average tropical cyclone maximum wind speed are likely to become more common in the future as the earth warms up. The recent Hurricane Sandy that struck the eastern United States could be an example of this.

MODIS image of Hurricane SANDY on 28th October 2012 (image courtesy NASA). According to many climate scientists, average tropical cyclone maximum wind speeds are likely to increase as the earth warms up.

What we did in our training during 2012

The training has partly focussed on the practical use of NWP products, especially from ensemble prediction systems. We’ve also covered reporting of severe weather events (which is required every 4 months), case studies (looking at a past weather event and trying to find out what happened and why) and verification (go here or here to see how we do it in NZ).

In all countries, either a half-day or full-day workshop devoted to Disaster Risk Reduction has been held. This has been a practical, interactive day with representation from the national meteorological service and many other agencies such as (but not restricted to) the National Disaster Management Offices, Civil Aviation, Maritime, Fire Service, Health departments and NGOs such as the Red Cross.

On this day we’ve run a mock-up severe weather exercise and asked participants to play their typical roles in responding to the expected weather impacts and formulating a plan of action.

Above: A mock-up severe weather exercise in Apia, May 2012 as part of the SWFDDP training.

Below are some images of the training in various countries over the last seven months:

Above: Group photo of facilitators and participants at the Niue SWFDDP Workshop, November 2012.

Above: Consulting Meteorologist Mark Schwarz discusses an exercise with staff at the Tuvalu Meteorological Service.

Above: Running a practical exercise at the SWFDDP DRR Workshop in Vanuatu, 2012.

Above: Manager RSMC Wellington, Steve Ready runs through a practical exercise on MetConnect Pacific with CIMS Directors Arona Ngari and staff.

Met staff in the Solomon Islands in a practical exercise

MetService WMO Manager James Lunny facilitates a practical exercise with staff at the Solomon Islands MetService as part of the SWFDDP training 2012.

And finally….we’ve been lucky to visit some unusual places and meet friendly people (from top to bottom: Funafuti lagoon, Tuvalu; Ministry football team, Solomon Islands; Schoolkids, Kiribati).

The mid-July northern low

Posted on August 4, 2009 by Jonathan Tunster

On the night of 17th July and early on the 18th, New Zealand was affected by a fast-moving and rapidly deepening depression originating in the north Tasman Sea. Sustained southwesterly winds of more than 60 knots were recorded in Colville Channel as the low passed by. Severe Weather Warnings were issued for wind in Coromandel/Great Barrier Island and rain in the eastern North Island.

This post will have a look at some of the reasons that the low deepened so rapidly, and whether the computer models did a good job of predicting its path. We’ll also see that weather is a lot more complicated than simply following what happens at the surface only.

: MODIS imagery of the low as it moved away from the country on Saturday 18th July

On Friday 17th July at 6PM, a low was analysed just west of Northland with a central pressure of about 997 hPa.

24 hours later the low lay a few hundred kilometres east of the North Island and had deepened to around 975 hPa. It continued deepening as it headed southeastwards – New Zealand was only affected by the first stages of the rapid development.

: Midday 17th July – a “slack low”

: Midnight 17th July – 12 hours later the low has quickly evolved

The rapid development of this low cannot be adequately explained by considering surface conditions only. Chris Webster here states “Perhaps we inadvertently reinforce a perception that our weather is 2-D by publishing lots of weather maps that are valid only at the Earth’s surface (e.g., see Weather Maps). Rarely do we show what’s happening higher up through the atmosphere or, more precisely, the troposphere —?the part of the atmosphere that contains our weather.”

Our low was in fact strongly influenced by what was happening at the highest reaches of the troposphere – the regions where jet streams tend to be present.

Upper levels

: ECMWF forecast for 3PM on the 17th. Green contours are surface isobars. Shaded regions are those of high upper level winds (black being the strongest).

This forecast chart is entirely computer generated, but shows why the low began organising itself and deepening from the Friday evening. The surface low pressure centre lies near the equator-ward entrance region of an upper level jet stream. This is a particularly favourable position for cyclogenesis (meaning development of a low). This is because it is an area of upper level divergence, which favours convergence at the surface and upward motion – both of which are conducive to developing depressions.
The other favourable position is the pole-ward exit region. Where these coincide (the exit of one jet, and the entrance of another), and where a low lies downstream from a sharpening upper trough then a recipe exists for “explosive cyclogenesis”:

Here are the comments made by senior forecasters regarding the development of this low:

“A sharpening upper trough over the Tasman Sea with double jet structure should provide strong upper level divergence for the low developing there.”

“Deepening caused by very strong divergence and vorticity associated with a broad very strong subtropical jet with winds reported 150-200 kt by aircraft”

…………………………………………………………………………………………………………………

The actual behaviour of the low

Analysis charts above, drawn every 1 hPa, show the actual track of the low as analysed by a senior meteorologist. Chart times are 6PM 17th (top left); 6AM 18th (top right); 9AM 18th (bottom left) and Noon 18th (bottom right).

Contrast the 6AM analysis with these forecast model prognoses for that same time:

UKMO Unified Model and ECMWF IFS model output - runtime midnight 15th July

Surface pressure contours. UKMO Unified Model (red) and ECMWF IFS model (blue) output for 6AM 18th July. Model runtime was midnight 15th July

Surface analysis for 6AM on the 18th July

There was considerable disagreement between models as to the track and depth of this low, and, as suggested by the analysis, no single model got it quite right.
Even though computer models are becoming ever more sophisticated, it’s important not to slavishly follow their predictions. As we have seen with this fairly brief event, it’s crucial – especially for high-impact weather events – to have professional meteorologists monitoring the situation and applying their judgement and conceptual knowledge of meteorology to the output of the models. This is something that is not likely to change in the future.

Meteorological Aspects of a Recent Trip to the Tararuas

Posted on May 19, 2009 by Jonathan Tunster

May 4th/5th 2009
I work as a public forecaster and my main tasks include writing regional, urban and mountain forecasts. I either work mornings (which start at 5:50AM and finish mid afternoon), or afternoons (which commence mid afternoon and run until 10:45PM).

A benefit of shiftwork is frequently having weekdays off, which earlier this month allowed me and Marine Forecaster Nicole Ranger to go tramping in the Tararua Forest Park. The state of the weather is an unavoidable facet of the outdoors, and so this first blog post will cover some of the meteorological aspects of this trip. Our destination was Mitre Flats Hut on the eastern side of the range, with an attempt on Mitre (highest peak in the Tararuas at 1571m) the following day.

A front was forecast to race up the South Island overnight and into the next day whilst forcing northwesterlies over the Tararuas. The good news was that this ought to clear up the the remaining cloud (from a dying southerly) on the eastern side of the range. The bad news was the threat of the winds being too strong on the exposed tops, and for Mitre itself to be clouded over.

Frost

The skies cleared on Monday night, resulting in a grass frost the next morning.

Image 1: Grass frost (Mitre Flats)

Image 2: Cumulus fragments showing wind direction aloft

Although it appeared to be a fine morning from the hut, sinister, scrappy cumulus fragments could be tracked coming off the main range as evidence that winds were now northwesterly aloft.

On cold nights, like the one we experienced, the surface layer of air “de-couples” from other layers of the atmosphere. This prevents stronger winds higher up from reaching the ground, and leads to the classic still, cold winter night. However, when the sun rises, the surface layer becomes destabilised, and winds can now mix down to the surface. This is just what happened on the frosty morning of May 5, and once the northwesterlies reached ground level the grass frost expired very quickly.

Wind

Heading up to Mitre, the wind was quite gusty in the bush, but it was much more so when we reached the exposed tops. The main peaks to our west had also clouded over in the northwest flow.

Image 3: Northwesterly winds ahead of the front were forced to rise over the Tararuas, forming clouds, and hence keeping the main spine of the range "clagged-in".

Image 4: Meanwhile, the Wairarapa is fine- thanks to the sheltering effect of the Tararua Range.

Unfortunately, we were not able to make it to Mitre this day.

Image 5: At about 1200m, before turning back. Back row - Altocumulus; Middle row - Marine Forecaster Nicole Ranger, Public Forecaster Jon Tunster; Front row - Holly the Dog

Despite our retreat, this was not the end of the interesting meteorology. The hint is in Image 5, with the quite impressive altocumulus behind us, which is aligned perpendicular to the northwesterly wind. This suggests significant “downslope winds”, which bring strong winds from aloft down to the surface in the form of gusts. When we returned to the hut (at about 400m above sea level), the wind was very gusty. A sharp-edged altocumulus to our east matched the theory of downslope winds.

A few days later, at the end of a shift, I found an archived satellite picture from that time of the day (1PM, May 5^th). The cloud signature was indicative of strong downslope winds in the lee of the Tararuas. Had I been on shift that day and seen such a satellite image, I would have expected this phenomena to be occurring- but being in the middle of it was much more fun…

Downslope Winds

front2

In the satellite picture, I’ve drawn the red arrow showing wind direction. Air in the northwesterly was forced up the hills, formed clouds and then descended on the lee side. This relatively dry, descending air was responsible for the strong wind gusts we experienced. Downstream, the air rises again, and forms the sharp edged line of altocumulus parallel to the Tararuas. It is this cloud formation which you can see in both the photograph from the ground and the satellite image from 36000km above. The clear area between the hills and the altocumulus is a variety of “foehn gap”- which is where I was standing when I took the photo. This cloud formation, seen either from the ground or on a satellite image, allows a good estimate to be made of where there may be especially strong and gusty downslope winds.

Jon Tunster, Public Forecaster May 17th 2009

MetService Blog

Author Archives: Jonathan Tunster