Atlantic Tropical Action 2014

[Ground Scouring]

The Euro is showing the same pattern over the next week that we just saw with the depression.

While we could see a spin up on the ITCZ, it won't have much of a future when it begins

gain latitude and ingest dry air.

For some reason, the 27km HWRF nest grid shows very low relative humidity and minimum precipitable water north of the ITCZ, yet as it detaches the low from the ITCZ, the low strengthens despite heading into an environment with a trade surge of low precipitable water. Given the sophistication of the HWRF, does anyone know why it's showing so much intensification?

[Kory]

For some reason, the 27km HWRF nest grid shows very low relative humidity and minimum precipitable water north of the ITCZ, yet as it detaches the low from the ITCZ, the low strengthens despite heading into an environment with a trade surge of low precipitable water. Given the sophistication of the HWRF, does anyone know why it's showing so much intensification?

I found the HWRF was always a bit on the bullish side when it came to strength.  

[Superstorm93]

Nice consistency from the GFS. 

 

Would like to see the ECM join the FIM8, FIM9, HWRF, CMC, NAVGEM, and UKMET 

 

Trough sits over the east for an ENTIRE week and eventually curves this out. 

 

[jconsor]

Regarding the ECMWF's record on forecasting TC genesis, this study is worth reading:

 

http://www.wunderground.com/blog/JeffMasters/genesis-of-new-atlantic-tropical-cyclones-which-model-should-you-trus

 

European ECMWF model: The model is reluctant to predict genesis, and misses many genesis events (it had only an 8% probability of detection in 2011.)... The model was highly prone to making late genesis forecasts.

 

American GFS model: The GFS model improved substantially in its genesis forecasts beginning in 2010, most likely due to a major model upgrade in 2010. The GFS is more aggressive at predicting genesis than the European model, and is less likely to miss a genesis event (22% probability of detection in 2011.) However, the incidence of false alarms was 32% in 2011, double what the European model had. 

 

UKMET model: The UKMET is more aggressive at predicting genesis than the European model, and is less likely to miss a genesis event (20% probability of detection in 2011.) The incidence of false alarms was 18% in 2011, similar to what the European model had. 

[bluewave]

Does anyone have the link to the new GFS? It seems to be better with the convective feedback

than the current version. Back in early June the GFS was going for a strong development

but the new version did better.

 

https://twitter.com/EricBlake12/status/476382638425780224/photo/1

 

[bluewave]

For some reason, the 27km HWRF nest grid shows very low relative humidity and minimum precipitable water north of the ITCZ, yet as it detaches the low from the ITCZ, the low strengthens despite heading into an environment with a trade surge of low precipitable water. Given the sophistication of the HWRF, does anyone know why it's showing so much intensification?

 

It could be that the GFS is picking up on some development potential here. But it looks like another system that may 

struggle to get much beyond TD or TS status due to the dry environment. Hopefully, it can bring some

rain to the islands which have been experiencing drought conditions.

 

 

St Lucia under ‘water emergency’ as desperate drought situation worsens

 

http://www.caribbean360.com/news/st-lucia-under-water-emergency-as-desperate-drought-situation-worsens

[phil882]

Regarding the ECMWF's record on forecasting TC genesis, this study is worth reading:

 

http://www.wunderground.com/blog/JeffMasters/genesis-of-new-atlantic-tropical-cyclones-which-model-should-you-trus

 

I'd prefer a peer reviewed journal article rather than a blog post from wunderground, but I've noticed as much just through observing the tropics in the Atlantic basin the last few years. The big ticket items the ECMWF does a pretty good job with, but its the weaker less substantial systems that it struggles to forecast genesis. I should also emphasize this is primarily at genesis, it tends to do a great job post genesis.

 

In any note, it seems to be picking up on our system on the 12 UTC run.

 

[dtk]

Does anyone have the link to the new GFS? It seems to be better with the convective feedback

than the current version. Back in early June the GFS was going for a strong development

but the new version did better.

 

https://twitter.com/EricBlake12/status/476382638425780224/photo/1

 

BpxzfBNIcAIVjMs.png

It is only being run by developers at EMC in quasi/near- real time in an unofficial capacity on the backup supercomputer.  There is currently machine maintenance ongoing in preparation for the phase 2 wcoss upgrade which means the parallel gfs isn't available (the backup machine is completely offline while they do some cabling work, I believe).  The experiment with the new gfs will be caught up again once the machine is available to non-production users.

 

When the experiment ("hw14") is running, some maps are plotted via the standard EMC verification webpage:

http://www.emc.ncep.noaa.gov/gmb/STATS_vsdb/

 

The plots are done for the 00z cycle.  Look for the menu under 2D maps that says 00z weather forecast maps.  There is an option for a set of graphics over the Atlantic basin.

 

It looks like the last set of graphics was done on the 24th before the development machine was taken away from users:

http://www.emc.ncep.noaa.gov/gmb/STATS_vsdb/allmodel/daily/2Dgfsfcst/arch/20140724/

 

I believe that the plan is still to implement sometime in Q1FY15, so there should be an official, real-time experiment managed by NCO up and running sometime soon (in the next month or two?)

[phil882]

 

http://www.americanwx.com/bb/index.php/topic/42483-atlantic-tropical-action-2014/?p=3010630

 

If we can finally get a pattern where the TUTTs lift north in prime time (latter August/earlier September), then the odds of a significant TC hitting the U.S. will be quite good, as virtually all the long-range guidance shows a strong, west-based Bermuda High (with a west-based -NAO and mean trough in the Midwest, as in 2004) during peak season. Over the next week, the SAL should also be decreasing as the flow off Mauritania shifts more to the N.

 

 

TUTTs and mid-level dry air are not mutually exclusive. There are two primary ways you can get widespread dry air into the MDR. The first is through vertical descent. When the MJO is active, normally one hemisphere has ample convection and a large-scale pattern that favors widespread ascent, where deep convection is produced as a byproduct. On the flip side, the other hemisphere struggles to produce convection because there is a suppressant lid due to large-scale vertical decent, which favors drying of the tropospheric column. Its the same reason why the Atlantic was so dry last week for TD#2, the active MJO was in the WPAC (and has been for some time if you look at the CPC RMM phase space trackers.

 

Secondly, mid-level dry air can be advected down by mid-latitude troughs on their western flanks. TUTTs can then continue to carry this dry air around the Atlantic while also increasing the vertical wind shear nearby. 

 

The real killer is if you get both of these processes working together, which appears to have been what caused 2013 to be so inactive. The MOT, which normally weakens and recedes poleward in August and September remained in place and in fact became stronger in some cases (e.g. TC Humberto moving due north in response to a strong TUTT around 30W). At the same time most of the TC activity was occurring in the WPAC due to an active MJO that seemly didn't budge (remember how there were 5 major TCs in the WPAC from August-October?).

 

I know there have been researchers (namely Dr. Gray and Dr. Klotzbach) that want to attribute 2013's inactivity to the PDO, and that might very well be a contributing piece, I think the evidence is also there that there are other factors at work driving active vs. inactive TC seasons.

 

The thing is, the source region of TUTTs haven't really been studied extensively. We know they originate in the mid-latitudes, but what types of patterns result in more prevalent TUTT activity vs. little TUTT activity aren't really well known. It wouldn't surprise me if they are tied to modes of the MJO and ENSO. And it goes beyond simply the frequency of TUTTs or mid-latitude troughs, but also the aspect ratio of these features. Thin troughs are a lot easier to split into TUTTs than thick troughs which are more likely to shear TCs and vertically stacked low-level cyclones apart.

 

When it comes down to it, we don't do a good job of predicting seasonal activity, because we don't have a handle on the synoptic changes in the tropics that occur from week to week, nor do we fully known what types of synoptic patterns in week 1 will lead to favorable conditions in the MDR in week 2. The MJO and ENSO are just two pieces to the puzzle, and knowing how the mid-latitudes interplay is critical. Its still an area ripe for research!

[bluewave]

It is only being run by developers at EMC in quasi/near- real time in an unofficial capacity on the backup supercomputer.  There is currently machine maintenance ongoing in preparation for the phase 2 wcoss upgrade which means the parallel gfs isn't available (the backup machine is completely offline while they do some cabling work, I believe).  The experiment with the new gfs will be caught up again once the machine is available to non-production users.

 

When the experiment ("hw14") is running, some maps are plotted via the standard EMC verification webpage:

http://www.emc.ncep.noaa.gov/gmb/STATS_vsdb/

 

The plots are done for the 00z cycle.  Look for the menu under 2D maps that says 00z weather forecast maps.  There is an option for a set of graphics over the Atlantic basin.

 

It looks like the last set of graphics was done on the 24th before the development machine was taken away from users:

http://www.emc.ncep.noaa.gov/gmb/STATS_vsdb/allmodel/daily/2Dgfsfcst/arch/20140724/

 

I believe that the plan is still to implement sometime in Q1FY15, so there should be an official, real-time experiment managed by NCO up and running sometime soon (in the next month or two?)

 

Thanks very much for the updated information on the GFS.

[Ground Scouring]

Regarding the ECMWF's record on forecasting TC genesis, this study is worth reading:

 

http://www.wunderground.com/blog/JeffMasters/genesis-of-new-atlantic-tropical-cyclones-which-model-should-you-trus

 

European ECMWF model: The model is reluctant to predict genesis, and misses many genesis events (it had only an 8% probability of detection in 2011.)... The model was highly prone to making late genesis forecasts.

 

American GFS model: The GFS model improved substantially in its genesis forecasts beginning in 2010, most likely due to a major model upgrade in 2010. The GFS is more aggressive at predicting genesis than the European model, and is less likely to miss a genesis event (22% probability of detection in 2011.) However, the incidence of false alarms was 32% in 2011, double what the European model had. 

 

UKMET model: The UKMET is more aggressive at predicting genesis than the European model, and is less likely to miss a genesis event (20% probability of detection in 2011.) The incidence of false alarms was 18% in 2011, similar to what the European model had. 

Which model had the best record in regard to TC intensity (both pre-genesis and post-genesis)?

 

TUTTs and mid-level dry air are not mutually exclusive. There are two primary ways you can get widespread dry air into the MDR. The first is through vertical descent. When the MJO is active, normally one hemisphere has ample convection and a large-scale pattern that favors widespread ascent, where deep convection is produced as a byproduct. On the flip side, the other hemisphere struggles to produce convection because there is a suppressant lid due to large-scale vertical decent, which favors drying of the tropospheric column. Its the same reason why the Atlantic was so dry last week for TD#2, the active MJO was in the WPAC (and has been for some time if you look at the CPC RMM phase space trackers.

 

Secondly, mid-level dry air can be advected down by mid-latitude troughs on their western flanks. TUTTs can then continue to carry this dry air around the Atlantic while also increasing the vertical wind shear nearby. 

 

The real killer is if you get both of these processes working together, which appears to have been what caused 2013 to be so inactive. The MOT, which normally weakens and recedes poleward in August and September remained in place and in fact became stronger in some cases (e.g. TC Humberto moving due north in response to a strong TUTT around 30W). At the same time most of the TC activity was occurring in the WPAC due to an active MJO that seemly didn't budge (remember how there were 5 major TCs in the WPAC from August-October?).

 

I know there have been researchers (namely Dr. Gray and Dr. Klotzbach) that want to attribute 2013's inactivity to the PDO, and that might very well be a contributing piece, I think the evidence is also there that there are other factors at work driving active vs. inactive TC seasons.

 

The thing is, the source region of TUTTs haven't really been studied extensively. We know they originate in the mid-latitudes, but what types of patterns result in more prevalent TUTT activity vs. little TUTT activity aren't really well known. It wouldn't surprise me if they are tied to modes of the MJO and ENSO. And it goes beyond simply the frequency of TUTTs or mid-latitude troughs, but also the aspect ratio of these features. Thin troughs are a lot easier to split into TUTTs than thick troughs which are more likely to shear TCs and vertically stacked low-level cyclones apart.

 

When it comes down to it, we don't do a good job of predicting seasonal activity, because we don't have a handle on the synoptic changes in the tropics that occur from week to week, nor do we fully known what types of synoptic patterns in week 1 will lead to favorable conditions in the MDR in week 2. The MJO and ENSO are just two pieces to the puzzle, and knowing how the mid-latitudes interplay is critical. Its still an area ripe for research!

Thanks very much for this fantastic response, Phil. In regard to 2013, I was also attempting to point out that I was referring not only to the thickness/depth of the MOT in that season, but also to its more conspicuous presence in the Atlantic since 2010. 2011 and 2012 in the Atlantic had many of the same problems that 2013 did. The more active MJO present in the WPAC would seem to fit with the warmer Indian Ocean/SW Pacific forcing active convection in that area. The same factors could also contribute to dryness in the Atlantic. I really believe that some large-scale climate factors are acting on the MOT, in terms of intraseasonal and interseasonal variation, and that climate change might well be one factor (though often dismissed) in terms of making deeper, more frequent MOTs in the Atlantic during peak season, adding in other factors like the modalities of the MJO, PDO, and ENSO. (Climate change obviously acts on those factors just as they act on weather/climate in their own right.)

[Superstorm93]

Really interested to see what the HWRF will do when that area near 30W is declared 93L. The 27km parent domain has been quite aggressive with this thing, showing a strong tropical storm approaching the islands (verbatim)

 

 

[phil882]

If future 93L develops, this would ultimately be a major coup for the GFS as it really was the first model to pick up on the feature.

 

Convection has significantly increased with the feature today, as the GFS suggested would happen. There is a nice pouch of moisture associated with the low-level vorticity, which is certainly more widespread than what could have been said with TD#2 last week.

 

[phil882]

The ongoing convection along the monsoon trough might be getting enhanced by a couple of upper-level cyclones which are located deep within the tropics, separated from the MOT. The upper-level diffluence produced by these two upper-level cyclones appears to be enhancing convection along 10N where our disturbance is located. This could help to explain why the GFS has been aggressive with convective development the next couple of days before TCG occurs.

 

The bigger question is what happens with the MOT, and embedded TUTT cells / PV streamers as we move into the medium range. The GFS wants to extend the upper-level anticyclone off of Africa further east by day 5 and push the MOT out of the way, while the ECMWF waits a few more days before that happens. 

[wxtrackercody]

Now an invest (93L) and now a cherry for the 5-day window.

 

 

 

 

*   ATLANTIC SHIPS INTENSITY FORECAST       *
                    * GOES AVAILABLE,       OHC AVAILABLE       *
                    *      INVEST  AL932014  07/28/14  06 UTC   *

TIME (HR)          0     6    12    18    24    36    48    60    72    84    96   108   120
V (KT) NO LAND    20    22    25    29    34    42    50    58    66    70    73    76    78
V (KT) LAND       20    22    25    29    34    42    50    58    66    70    73    76    78
V (KT) LGE mod    20    21    22    24    26    31    36    43    51    61    70    75    77
Storm Type      TROP  TROP  TROP  TROP  TROP  TROP  TROP  TROP  TROP  TROP  TROP  TROP  TROP

SHEAR (KT)         3     4     4     6     7    11     9    10     2     2    10    10    13
SHEAR ADJ (KT)     4     6     7     6     9     8     6     4     4     1    -1    -3    -1
SHEAR DIR        113    78    67    61    28    32    35    53   348   253   222   185   192
SST ©         27.5  27.4  27.2  27.0  26.8  26.5  26.3  26.4  26.8  27.2  27.3  27.3  27.5
POT. INT. (KT)   131   130   128   125   123   121   119   119   124   128   129   129   131
ADJ. POT. INT.   132   131   129   126   124   122   119   120   124   128   128   128   129
200 MB T ©   -54.1 -54.1 -53.8 -54.4 -54.7 -53.7 -54.1 -53.7 -54.2 -54.0 -54.3 -53.4 -53.7
TH_E DEV ©       4     4     5     5     5     5     6     6     8     8     9    10    11
700-500 MB RH     62    62    62    62    62    56    54    51    51    50    48    45    45
GFS VTEX (KT)      0  LOST  LOST  LOST  LOST  LOST  LOST  LOST  LOST  LOST  LOST  LOST  LOST
850 MB ENV VOR    -8     5    19    31    35    30    27    14    20    18    13     2    -3
200 MB DIV        43    34    53    88    80    35    46   -25   -13     7    42    41    78
700-850 TADV       0     0    -1     0    -3   -13   -10   -10    -7    -9    -2    -5     1
LAND (KM)       1547  1677  1649  1601  1565  1525  1270  1071   910   832   791   558   446
LAT (DEG N)     10.3  10.4  10.5  10.6  10.7  11.0  11.3  11.7  12.1  12.7  13.1  13.7  14.5
LONG(DEG W)     31.0  32.3  33.6  35.0  36.3  39.2  42.3  45.4  48.4  51.3  54.0  56.8  59.4
STM SPEED (KT)    12    13    13    13    14    15    15    15    15    14    13    14    13
HEAT CONTENT      16    15    13    12     8     2     3    14    15    22    26    24    27

[Sunny and Warm]

The early guidance really likes 93L:

 

[GaWx]

Regarding 93L, despite the overall guidance developing it, the Euro still does little with it. Opinion about the pretty lackluster Euro as compared to other models for 93L?

[Srain]

There continues to be a lot of dust to the N of 93L. Conditions appear marginal favorable for some development as long as this disturbance remains attached to the ITCZ where a rather low shear environment and instability are some what positive for TC genesis. I suspect the next Easterly wave exiting Africa may offer a slightly better environment over next couple of days as it continues W near 10N.

 

We will see if showers/storms associated with 93L continue to persist as it travels W at fairly low latitude over the next couple of days around 10 to 15 miles per hour. I would like to see convection remain somewhat in tact for a day or two before jumping onboard a solution similar to what the GFS is suggesting.

[Kory]

Regarding 93L, despite the overall guidance developing it, the Euro still does little with it. Opinion about the pretty lackluster Euro as compared to other models for 93L?

Looks like forward speed might be the key.  The Euro is faster with it moving westward, which keeps it weaker and thus further south more towards the shear.  Looking at the Euro 500 mb pattern, it shows something similar to the GFS with a recurve out to sea (probably by 75 W), although I do suspect a shift to the left from the GFS as it's usually placed to far to the right at this point in the game.  

[RaleighWx]

The Euro has struggled some last few seasons with the incipient stages of these deep atlantic waves.

 

With most medium/long range guidance showing a trough on east coast, seems to me the key is how far south it stays over the next 5-7 days. The trough could weaken and lift out 8-10 days from now, if Invest93(Bertha) is far enough south it could be a threat down the road, however if it gets further north like the GFS/GGEM show re-curve chances would be high.

 

Latest intensity models

[Ground Scouring]

Over the next two days (48hr), the ECMWF has 700-mb wind vectors more out of the NE, whereas the GFS has vectors more out of the ENE/E. This makes a big difference because the ECMWF solution allows more dry air/SAL to advect into the circulation of 93L, which is a big reason why the ECMWF loses a closed low by day three. The ECMWF builds less of a shortwave ridge wedging between 93L and the MOT low near the Azores, which means both weaker low-level steering (slower short-term movement than on the GFS) and a weaker 93L. Both these factors offset each other somewhat, but the big thing to take from satellite and model data is that the GFS is likely developing 93L too quickly (a common phenomenon when it picks up on development), showing a depression 24hr out. Note the low-level stratocumulus currently advecting into the circulation from the NE, which explains why convection on the east side of the broad circulation is meager. This trend should continue over the next two days. A blend of the ECMWF and GFS data would suggest development about a day later than shown on the GFS. But to get a sense of where 93L is heading in the medium/long range, we need to see a closed low develop first.

 

Edit: 93L

[phil882]

Over the next two days (48hr), the ECMWF has 700-mb wind vectors more out of the NE, whereas the GFS has vectors more out of the ENE/E. This makes a big difference because the ECMWF solution allows more dry air/SAL to advect into the circulation of 93L, which is a big reason why the ECMWF loses a closed low by day three. The ECMWF builds less of a shortwave ridge wedging between 93L and the MOT low near the Azores, which means both weaker low-level steering (slower short-term movement than on the GFS) and a weaker 93L. Both these factors offset each other somewhat, but the big thing to take from satellite and model data is that the GFS is likely developing 93L too quickly (a common phenomenon when it picks up on development), showing a depression 24hr out. Note the low-level stratocumulus currently advecting into the circulation from the NE, which explains why convection on the east side of the broad circulation is meager. This trend should continue over the next two days. A blend of the ECMWF and GFS data would suggest development about a day later than shown on the GFS. But to get a sense of where 93L is heading in the medium/long range, we need to see a closed low develop first.

 

Edit: 93L

 

The ECMWF in fact has 93L moving quicker to the west than the GFS does before it opens back up into a wave. I don't disagree with your thinking Re: SAL. If stronger northeasterly flow is advected into 93L as a result of slower forward motion it may struggle to produce convectively. However, I don't see any evidence of the system slowing down right now, and the low-level ridge isn't significantly weaker than normal, with most of the models having 93L moving as a pretty nice clip (around 15 kt wnw motion). 

 

I'm in the camp thinking that as long as shear remains low, the SAL in 93L's proximity will not be advected into the circulation. This system has a much broader moisture pouch than TD#2 owing to the fact that this is a larger disturbance. The system might have to contend with increasing westerly vertical wind shear as it approaches the lesser antilles, but as for now, I think things look pretty good for TCG. 

[negative-nao]

You can see on the infrared that the clouds are starting to flare out.

 

http://www.ssd.noaa.gov/PS/TROP/floaters/93L/imagery/rb_lalo-animated.gif

[icebreaker5221]

12Z ECMWF coming in with a much better defined vortex.  Still on the weak end of the guidance relative to GFS/CMC/NAVGEM, but that was to be expected.  I'm thinking 93L has a good shot at becoming a TD within 24-36 h. 

 

Euro also takes it much further south with the more easterly low-level steering flow (not to mention a weaker vortex equates to less northward Beta-drift), with a track south or Puerto Rico / towards southern Hispanola. 

[phil882]

12Z ECMWF coming in with a much better defined vortex.  Still on the weak end of the guidance relative to GFS/CMC/NAVGEM, but that was to be expected.  I'm thinking 93L has a good shot at becoming a TD within 24-36 h. 

 

Euro also takes it much further south with the more easterly low-level steering flow (not to mention a weaker vortex equates to less northward Beta-drift), with a track south or Puerto Rico / towards southern Hispanola. 

 

GFS 96 Hour Forecast (200-hPa winds)

 

ECMWF 96 Hour Forecast (200-hPa winds)

 

 

The ECMWF also has a vastly different upper-tropospheric pattern beyond 72 hours in comparison to the GFS. The GFS has a strong TC with active convection driving upper-tropospheric outflow, which essentially pushes the MOT out of the way and allows the TC to continue to maintain or intensify under the weaker upper-level flow overhead.

 

The ECMWF seems to be have much weaker convection (possibly due to dry air intrusion?) that results in more limited upper-tropospheric outflow. As a response, the ECMWF is weaker and more sheared since weaker convection isn't mitigating the light westerly upper-tropospheric flow provided by the MOT in its path. The system is still trackable until it gets to Hispaniola, where it then promptly gets shredded apart. 

 

Which one is right? I know we have been poo pooing the ECMWF solution for a while now, but once the system becomes a coherent TC, the ECMWF has a superior convective scheme that is better than the GFS. I get worried when I see the system failing to develop a good PV tower before it reaches the less favorable upper-tropospheric flow. 

 

I think the fly in the ointment is the dry air (SAL?) to the north and whether or not dry air intrusion erodes away at the nice moisture plume over the system currently. The GFS obviously doesn't think that happens, but the ECWMF has other ideas.

 

Either way it's shaping up to be a fascinating forecast. These are the type of ambiguous cases that I really enjoy trying to forecast!

[Kory]

Interesting note...A lot of the GFS ensembles are west of it's operational run.

[phil882]

Will probably see those probabilities go up at 8pm given the latest ASCAT and Microwave passes

 

 

[wxtrackercody]

Will probably see those probabilities go up at 8pm given the latest ASCAT and Microwave passes

 

 

Definitely organizing at a quick pace this afternoon. The bulk of the convectively-coupled kelvin wave should be passing over the wave right now.

 

[Chaser25973]

If this storm passes south of PR, what are the analogs? Could it split the pylons between Mexico and Cuba and be a threat to the Gulf Coast?

[wxtrackercody]

If this storm passes south of PR, what are the analogs? Could it split the pylons between Mexico and Cuba and be a threat to the Gulf Coast?

Based on everything I'm seeing, there's really two options here. The first, more likely solution (IMO) is that this gains strength as it heads towards the extreme northern Leeward Islands and eventually recurves out to sea east of the United States. The second, which is the less likely but the solution that the ECMWF offers up, is that this remains weak and enters the Caribbean Sea, where it's ripped apart by strong upper-level winds and/or a fast surface trade wind flow and/or the mountainous terrain of the Greater Antilles. I don't see how this could pose a threat to the Gulf Coast.