19W develops

September 24th, 2008

As of 12Z on the 24th, 19W was analyzed near 12.6N 136.3E according to JTWC. It is moving 295 degrees at 10 kts with max sustained winds at 045 kts and gusts to 55 kts. The current Satcon (as of 07Z) analyzes its intensity at 46 kts. The 12:30Z ADT gives an intensity of 47 kts.

The attached IR/upper level AMV and analyzed divergence plot shows the organized circulation, well defined outflow and divergence:

The attached MIMIC animation shows some of the convection organizing to its south as it is trying to develop an eyewall:

The shear over the center is still pretty low (around 5kts) although the surrounding shear is still fairly large (nearly 30 to 40 kts to the southwest).

So what about the future? JTWC has it intensifying by 10 kts every 12 hours or so, with intensities of 120 kts by 12Z on the 19th.

In terms of track, the models are in reasonable agreement, bringing 19W to the northwest close to Taiwan by 00Z on the 29th. JTWC has it to the southeast of Taiwan at that time: closer to the Met Office and NOGAPS solution. The GFS and ECMWF have it much closer to Taiwan.

Over the long term, the ECMWF brings 19W into the Chinese mainland as it moves westward. It also develops TCS048 over the long term, bringing along a similar track as 19W. It’s done fairly well with these genesis cases, so it is worth watching.

The ECMWF ensemble probability predicts strong vorticity along its deterministic track to be quite likely. It is a little less certain over the long range.

Links:

ECMWF, Met Office, NOGAPS, GFS 850 hpa vorticity.

ECMWF 850 vorticity probabilities.

TCS047 develops

September 23rd, 2008

The following two images suggest that TCS047 has sustained winds near 25 kts:

Surface observations and IR Image for TCS047. The ship VNVR is reporting a wind speed of 25kts

Surface observations and IR Image for TCS047. The ship VNVR is reporting a wind speed of 25kts

Quikscat Analysis and IR images for TCS047. The analysis shows a clear surface circulation

Quikscat Analysis and IR images for TCS047. The analysis shows a clear surface circulation

The divergence is fairly strong over some of the larger convective areas, although it hasn’t yet developed an anticyclonic flow aloft:

Although the shear is strong around the storm, it has weakened a great deal over the storm center:

What is driving this development? Well, TCS047 has been under the influence of a weak upper-level low to its north. In the attached animation of upper-level AMVs you can see the weak low travel slowly northwest, perhaps allowing TCS047 better outflow and development. In this animation, 47 is at about 10N 140E:

22SEP2008-upperwinds_anim.gif

And what about the model forecasts:

In terms of track, the models are remarkably similar, even through 120 hours.

The GFS and ECMWF bring this, by then, well developed storm just to the east of Taiwan.

The Met Office model and the NOGAPS bring it farther south, with the Met Office model a bit to the west.

All of the major models have some form of intensification.

What’s going on in the upper-levels?

The ECMWF 200 hPa forecast quickly (00Z on the 25th) erodes the upper level low and develops anticylonic flow over TCS047 as it tracks northwestward, steered by an upper level anticyclone to the west. It eventually stalls it out as this ridge moves out to the east. Hopefully this will remain around long enough for some interesting flights.

ECMWF takes a turn to the south on TCS047

September 22nd, 2008

The 12Z EC run is very different than the 00Z run. The 00Z run had TCS043 develop and then recurve and shear out. The 12Z run has TCS043 developing, and then turning directly west towards Taiwan.

It seems to be coming in line in track with some of the other models. It will be worth watching as the ECWMF has been very consistent throughout this experiment.

TCS047: To develop or not to develop

September 22nd, 2008

TCS047 is the next TCS invest of interest.

Attached is the 12Z IR image, Scatt data and surface observations available nearby:

IR Image, surface, ship, buoy reports, Scatt data and low-level AMVs for potentially developing system TCS047.

IR Image, surface, ship, buoy reports, Scatt data and low-level AMVs for potentially developing system TCS047.

TCS047 is not very well developed or organized. Most of the surface winds above are easterly, although there are hints of southwest winds from the scatt images and a few ship reports.

Although there is some weak divergence aloft and a low level vorticity center. It’s just to the south of an upper low, and that may be causing the shear surrounding the invest. The shear is fairly weak over the invest center, but it is quite strong to the west and north.

See the attached divergence image, and vorticity/shear image:

Upper level divergence and AMVS (top) and deep shear and low-level vorticity (bottom). AMVs and analyses are valid 16Z on 22 September. The IR images are valid at 11 Z.

Upper level divergence and AMVS (top) and deep shear and low-level vorticity (bottom). AMVs and analyses are valid 16Z on 22 September. The IR images are valid at 11 Z.

Of course the big question is when and if this system will develop.

The Met Office model weakens it through 00Z on the 24th and then slowly strengthens it, moving it to the northwest at about 18N 130E by 00Z on the 27th.

The ECMWF model analyzes the storm fairly well. It’s 12 hour forecast matches pretty well with the cloud pattern seen above. It develops this storm over the next few hours and begins to recurve it by around 12Z on the 26nd. By the end of the forecast period: October 02, the ECMWF has a very sheared system located to the south of Japan.

The GFS doesn’t really develop it at all past the short term. As it moves it north and west it kills it off, merging it with some other convective elements up near the Phillipeans. The 12Z run seems to be developing it a bit more actually.  The convection seems a bit more organized. Is this a trend?

It seems there are some differences in the long term shear. See the 00z run here and the 12z shear/850 vorticity here. Look at the forecasts at 00Z on the 26th. The 12Z has a more organized system that is farther to the west than the 00Z run. It’s also under an area of lower shear. Perhaps this is encouraging it’s development.

The NOGAPS that never made it a disturbance it didn’t like, develops it steadily throughout the forecast period and brings it up east of the Phillipeans by 00Z on the 28th.

Taking a quick look at the upper levels in the two models as shown here implies that the 12Z forecaster has a weaker upper level jet, perhaps contributing to the weaker easterly flow.

Hagupit: 12Z September, 22, 2008

September 22nd, 2008

From JTWC as of 1200 Hagupit is near 19.5N 120.9E, moving 280 degrees at 11 KTS. It’s at 100 kt gusts to 125 kts. The satcon intensity estimate is attached below:

08Z Satcon analysis on September 22nd, 2008 for Typhoon Hagupit.

08Z Satcon analysis on September 22nd, 2008 for Typhoon Hagupit.

As you can see Hagupit has been intensifying steadily over the past few days. The ADT seems to be above the JTWC, possibly due to the well developed eye and structure. When the next AMSU pass comes through, perhaps SATCON will catch up. Unfortunately, there will no aircraft penetration for this storm due to some mechanical problems. Perhaps some of the surface obs and buoys will give some clues.

The attached MIMIC animation shows how the convection has been organizing over the past few days as Hagupit moved north past the Phillipeans:

Mimic animation showing Hagupit's recent intensification

Mimic animation showing Hagupit's recent intensification

Hagupit is in a region of low shear and reasonable outflow to the west. It is forecasted to weaken slightly before landfall (about 00Z on the 24th) as seen in the JTWC forecast graphic.

The attached IR image and upper-level AMVs shown below show the strong, but focused outflow to the west:

According to JTWC, the strong ridge to Hagupit’s north has restricted some northerly outflow. One can’t see a strong anti-cyclone over or just outside of the storm.

The models are in fairly good agreement with the track, bringing it just south of Hong Kong by about 00Z on the 24th. There is a little variation: NOGAPS is a bit to the east of the suite of other models.

It’s under fairly weak shear now. JTWC is expecting the shear to increase slightly, thus causing its slight weakening before landfall.

Tropical Storm Hagupit is born!

September 19th, 2008

On 12Z the JTWC analyzed Hagupit at 14.0N 133.0E moving  265 degrees at 11 kts, with sustained winds of 40kts, gusts to 50 kts.

Tha 08Z AMSU detected an intensity of 42 kts (the satcon assigned it to 40 kts). The 17:30Z ADT gave it a CI of 3.1, corresponding to 47 kts. Presumably a future AMSU pass will give it a higher satcon intensity.

The mimic animation, shown below is interesting:

As the animation shows, the convection seems to be developing on the southwest side of Hagupit. Although it’s development is slow, the convection is becoming more widespread.

The attached IR image and ASCAT analysis shows a definitive closed circulation:

MTSAT IR image and ASCAT surface analysis for Hagupit. The analysis confirms a closed surface circulation.

MTSAT IR image and ASCAT surface analysis for Hagupit. The analysis confirms a closed surface circulation.

The shear has lightened up slightly over the storm’s center, probably precipitating some of the recent intensification. But, it is still limiting development to it’s southern side, as seen in the divergence and shear plots shown below:

MTSAT IR images, upper level AMVs and analyzed divergence. Valid 18Z September 9th, 2008.

MTSAT IR images, upper level AMVs and analyzed divergence. Valid 18Z September 19th, 2008.

MTSAT IR Image, and analyzed deep layer shear. Valid 18Z September 19th, 2008.

MTSAT IR Image, and analyzed deep layer shear. Valid 18Z September 19th, 2008.

So what is in Hagupit’s future?

JTWC has it intensifying steadily throughout the forecast period, bringing it to 105 kts by 12Z on the 24th.

Their track, and various model tracks are shown with the IR image below:

The models are in reasonably agreement through 36 hours, bringing it steadily northwestward. After that, they begin to diverge. The NOGAPS model recurves Hagupit, bringing it east of Taiwan and up toward Okinawa.  The Met Office model and ECMWF take it just north of the Phillipeans and then west into China.  The GFS takes the middle course, bringing Hagupit across Taiwan and into the Chinese mainland much farther north.

The track seems to depend on an upper-level trough and ridge to Hagupit’s north by about 00Z on the 22nd. The Met Office and ECWMF place Hagupit west of the trough, and under the control of an upper level ridge over China. The NOGAPS model moves the trough just to the west of Hagupit. The ridge over China is farther to the west than the Met Office and ECWMF, causing the storm to recurve between the trough and the subtropical ridge to its east. The GFS has Hagupit nearly direcly under the trough, almost equi-distance between the two ridges. This is definitely the storm to watch, and may be a potential future flight.

Links to 200 hPa streamlines:

NOGAPS,GFS,ECMWF,Met Office

Sinlaku: Forecast and future

September 19th, 2008

By T+48 (00Z on the 19th) most of the models are fairly similar. All of the models except the EC seem to have maintained symmetry by that time. Almost all of the models have Sinlaku near about 150E, although the Met Office model is a bit slower with its eastward progression. The NOGAPS and the GFS are a bit farther north than the other models as well. After ET sets in, the models move the former Sinlaku at slightly different paces. NOGAPS seems a bit faster by +108, bringing Sinlaku just near the international date line.

The other three models have Sinlaku at about 170 E…moving it along a little bit slower. The EC model is the only one to keep Sinlaku south of 40 N.

Links: NOGAPS GFS ECMWF Met Office

What about the phase diagrams?

The GFS diagram already diagnoses an assymetric vortex. It  transitions Sinlaku to a cold core vortex just before the 21st, and then makes it symmetric again by 18Z on the 23rd.

The NOGAPS diagram also transitions to a cold core by the 21st, starting from an assymetric vortex. It maintains this fully extra-tropical storm throughout the forecast period.

The 12Z Met Office model diagram has Sinlaku just on the edge of being a cold core system. Towards the end of it’s forecast period, it also makes it more symmetric.

The plan is to continue to fly Sinlaku from 00Z (C-130) and 02Z (P-3) on the 20th. Hopefully this will capture and measure the continued transition.

Sinlaku is a Typhoon once again…and then back to a Tropical Storm

September 19th, 2008

As of 00Z on the 19th, Sinlaku was upgraded once again to a Typhoon by JTWC. Then, by 12Z on the 19th it was downgraded to a tropical storm

This Mimic animation shows Sinlaku’s eyewall development with convection building to it’s south, and then subsequent weakening:

MIMIC animation of Sinlaku from 12Z on the 18th to 12Z on the 19th. Notice how convection develops on its southwest side as it intensifies.

MIMIC animation of Sinlaku from 12Z on the 18th to 12Z on the 19th. Notice how convection develops on its southwest side as it intensifies.

Sinlaku is still under the duel effects of shear and strong upper-level outflow from the midlatitude westerlies. Both the divergence and shear plots are attached below:

IR MTSAT Image, upper-level AMVs and analyzed divergence. Valid 13Z on September 19th, 2008.
IR MTSAT Image, upper-level AMVs and analyzed divergence. Valid 13Z on September 19th, 2008
MTSAT IR Image and upper-level shear. Valid 13Z September 19th, 2008.

MTSAT IR Image and upper-level shear. Valid 13Z September 19th, 2008.

Again, Sinlaku’s motion is aligned fairly well along the shear axis, perhaps limiting it’s negative impact on its intensity somewhat. The enhanced outflow from the upper-levels and relatively high SSTs are competing with the shear. The shear will likely win out and Sinlaku will weaken.

Satcon shows the weakening as well, with the ADT producing much lower intensities than previous satcon analyses:

Satcon for Sinlaku. The ADT is currently producing weaker estimates than previous satcon estimates.

Satcon for Sinlaku. The ADT is currently producing weaker estimates than previous satcon estimates.

Yesterday and today the P-3 and C-130 will focus on Sinlaku and try and capture it’s ET. This is first time that extra-tropical transition has been captured by recon in the west pacific.

What about the forecast?

They will be the focus of subsequent blogs.

Sinlaku: 13Z, September 18th, 2008

September 18th, 2008

Sinlaku is still churning away. It was briefly upgraded to a Typhoon at 00Z on the 18th, then downgraded back to a Tropical Cyclone.  According to JTWC, as of 12Z on the 18th, it was located near 31.0N 131.5E, moving 060 degrees at 09 kts. They analyze it at 55 kts with gusts to 70 kts.

Attached below is the standard IR/AMV/divergence plot valid 13Z on September 18th:

Although Sinlaku has continued outflow to the northeast, it’s flow is very well contained on the west side as it gets caught up in the midlatitude westerlies. The shear over Sinlaku, shown below, is very strong from the west:

The satcon analysis below shows Sinlaku’s recent intensity trend:

You can see the brief peak in intensity from JTWC (black curve) and the slight upward trend from SATCON and the ADT. Reading their prognostic reasoning, it appears JTWC re-intensified the storm based on recon data from TPARC.

Sinlaku may have re-intensified partially because its center has remained over water. This may allow it to maintain its tropical characteristics a bit longer than if it had passed over land.

As it begins ET, we will focus a bit on the tilt of its vortex. Currently, its low level center is a bit displaced to the southwest from the maximum convection as seen in the attached 85 GhZ and divergence plot:

The displacement isn’t huge, but the center (shown by the hurricane symbol, and the slightly circular pattern in the microwave image) is slightly over from the more intense microwave images and the divergence maximum. As ET happens, we’d expect more a vortex title. We’ll explore other ways to show this.

In the short term, the models are fairly consistent in bringing Sinlaku just south of the Japanese islands. There are some slight differences in terms of time and storm size, but the broad picture is fairly similar.

After Sinlaku leaves Japan, the models take a few different tacks with it’s track. The Met Office model and NOGAPS both bring Sinlaku up fairly far north and intensifying it, threathening the Aleutian Islands.

The ECMWF model merges Sinlaku with another midlatitude system, eventually moving toward Alaska.

The GFS takes a middle path: bringing Sinlaku up the Aleutians, but at a slower pace, and with less intensity than the Met Office.

It’s quite easy to see how uncertainties at this stage of forecasting can easily affect long-range forecasting downstream on the US coast.

And the phase diagrams? The GFS diagram begins to develop an assymetric core by the 19th, but does not really become cold core until the 24th of September!

The NOGAPS is faster: generating the assymetric core right away with Sinlaku becoming cold core just before the 20th.

The Met Office model diagram keeps it warm core throughout the shown forecast period.

How will Sinlaku interact with the midlatitute jets and outflow? When will it become extratropical? How strong will it get? These are questions that will need further investigation.

TCS043

September 17th, 2008

Attached is an IR animation for TCS043:

Taking a look at a two day animation (focus on TCS043 at about 15N 142E) you can see dirunal variation, as TCS043 strengthens around 12Z (22:00 Guam time) and weakens around 00Z (10 am Guam time). The short animation above is from 10Z - 15Z: near the convective maximum.

At this point, however, shear is the primary factor influencing TCS043:

The above plots show the shear (top) and divergence (bottom) over TCS043. The shear is consitently pushing the convection south. You can see this with the IR images, and with the location of maximum divergence. It’s unlikely to develop while under such strong shear.

The models are in two camps with this invest. The NOGAPS and GFS model have TCS043 developing and then recurving around (NOGAPS) and past (GFS) 140 E longitude. The 12Z ECWMF still has TCS043 moving west and interacting with the monsoon near the Philippians. Unlike the previous run, this run suggest hints of recurvature towards the end of the forecast period. The Met Office model, always the most reluctant developer, brings TCS043 westward and has a weaker interaction with that Phillipean convection.

The key to the recurve/no recurve difference seems to stem from a trough coming off of the China coast from the 20th-21st. The models with a deeper trough seem to be recurving the storm. The models without it seem to allow TCS043 to progress westward. We’ll have to keep an eye on this track difference in future runs.