2020 Indicators: SST's / SAL / MSLP / Steering / Shear / Instability / Sat Images

[AnnularCane]

It’ll be interesting to see the evolution of SST’s in the Atlantic over the next few months. Right now it’s a blend of a +AMO and a -AMO.

https://twitter.com/cyclonicwx/status/1263877121539178496

I personally think ECMWF + UKMET Superblend is onto something considered the current Atlantic SST configuration. I’m thinking most of not all waves that develop in the Tropical Atlantic will be low riders due to where the warmth is displaced south.

https://twitter.com/bennollweather/status/1263850857369493505

The output strongly suggests that threats to the mainland U.S., if any, would be largely homegrown in nature. Note that the northwestern Caribbean and Eastern Seaboard are largely “dry,” with most systems remaining well to the south of the Greater Antilles. Only the Gulf of Mexico features above-average precipitation in proximity to the mainland U.S. In terms of wind, this is potentially good news, since it suggests that any hurricane impacts to the U.S. would be weak. Going back to 1851, more than three-fourths of all major hurricanes to hit the U.S. developed in the MDR and were thus not homegrown systems. Most homegrown hurricanes tend to be weak (Cat-1/-2) at landfall in the U.S. The UK/EC blend indicates that the MDR-related activity will stay well to the south of the mainland U.S. in ASO of 2020.

I said it before and I guess I will have to say it again. Never underestimate the Gulf. Offhand I can think of quite a number of storms that blew up there and sometimes leading people to have to leave in a hurry.

[Hurricaneman]

I personally think ECMWF + UKMET Superblend is onto something considered the current Atlantic SST configuration. I’m thinking most of not all waves that develop in the Tropical Atlantic will be low riders due to where the warmth is displaced south.

https://twitter.com/bennollweather/status/1263850857369493505

The output strongly suggests that threats to the mainland U.S., if any, would be largely homegrown in nature. Note that the northwestern Caribbean and Eastern Seaboard are largely “dry,” with most systems remaining well to the south of the Greater Antilles. Only the Gulf of Mexico features above-average precipitation in proximity to the mainland U.S. In terms of wind, this is potentially good news, since it suggests that any hurricane impacts to the U.S. would be weak. Going back to 1851, more than three-fourths of all major hurricanes to hit the U.S. developed in the MDR and were thus not homegrown systems. Most homegrown hurricanes tend to be weak (Cat-1/-2) at landfall in the U.S. The UK/EC blend indicates that the MDR-related activity will stay well to the south of the mainland U.S. in ASO of 2020.

I said it before and I guess I will have to say it again. Never underestimate the Gulf. Offhand I can think of quite a number of storms that blew up there and sometimes leading people to have to leave in a hurry.

Or homegrown systems period as here are a few examples

Carol 1954 became a ts in the Bahamas
Bob 1991 became a Ts in the Bahamas
Belle 1976 Became a Ts in the Bahamas
Elena 1985( became a tropical storm in Cuba)
Opal 1995 formed in the BOC
Hermine 2016 formed 200mi west of Florida
Erin 1995 Formed near the Bahamas
Rita 2005 formed near the Bahamas
Katrina 2005 formed near the Bahamas
Arthur 2014 formed in the Bahamas
Harvey 2017 Became its strongest in the GOM
I’ll even throw in the Christmas Eve nor’easter of 1994 which formed near Florida and I do believe this system was a hurricane of 85mph before it left the Gulfstream

And some near misses from home brew like

Joaquin 2015 formed between Bermuda an North Carolina

So based on these Home brew need to be watched just as much as any And some were tropical waves until the Bahamas

[ScottNAtlanta]

I personally think ECMWF + UKMET Superblend is onto something considered the current Atlantic SST configuration. I’m thinking most of not all waves that develop in the Tropical Atlantic will be low riders due to where the warmth is displaced south.

https://twitter.com/bennollweather/status/1263850857369493505

The output strongly suggests that threats to the mainland U.S., if any, would be largely homegrown in nature. Note that the northwestern Caribbean and Eastern Seaboard are largely “dry,” with most systems remaining well to the south of the Greater Antilles. Only the Gulf of Mexico features above-average precipitation in proximity to the mainland U.S. In terms of wind, this is potentially good news, since it suggests that any hurricane impacts to the U.S. would be weak. Going back to 1851, more than three-fourths of all major hurricanes to hit the U.S. developed in the MDR and were thus not homegrown systems. Most homegrown hurricanes tend to be weak (Cat-1/-2) at landfall in the U.S. The UK/EC blend indicates that the MDR-related activity will stay well to the south of the mainland U.S. in ASO of 2020.

I said it before and I guess I will have to say it again. Never underestimate the Gulf. Offhand I can think of quite a number of storms that blew up there and sometimes leading people to have to leave in a hurry.

Opal '95 comes to mind as one where we went to bed looking at a minimal hurricane and woke up with a Major barreling towards the panhandle.

[AnnularCane]

The output strongly suggests that threats to the mainland U.S., if any, would be largely homegrown in nature. Note that the northwestern Caribbean and Eastern Seaboard are largely “dry,” with most systems remaining well to the south of the Greater Antilles. Only the Gulf of Mexico features above-average precipitation in proximity to the mainland U.S. In terms of wind, this is potentially good news, since it suggests that any hurricane impacts to the U.S. would be weak. Going back to 1851, more than three-fourths of all major hurricanes to hit the U.S. developed in the MDR and were thus not homegrown systems. Most homegrown hurricanes tend to be weak (Cat-1/-2) at landfall in the U.S. The UK/EC blend indicates that the MDR-related activity will stay well to the south of the mainland U.S. in ASO of 2020.

I said it before and I guess I will have to say it again. Never underestimate the Gulf. Offhand I can think of quite a number of storms that blew up there and sometimes leading people to have to leave in a hurry.

Opal '95 comes to mind as one where we went to bed looking at a minimal hurricane and woke up with a Major barreling towards the panhandle.

I still remember the shock at waking up to a 150 mph cane barreling toward land. I was in NOLA then, fortunately far enough west of the expected landfall, but it could have easily been us were it not for that trough being where it was.

[Shell Mound]

Didn't NOAA just yesterday state that it is IMPOSSIBLE to forecast storm tracks close in as even just two weeks?

These precipitation plumes are sometimes a good tool to use as an indicator. The last 2-3 seasons it did pretty well indicating overall storm clusterings and tracks.

Not really. TCs are small scale features that cannot accurately be picked up by monthly averaged precipitation anomaly maps. The last two seasons the Euro showed a bone dry Africa and MDR that did not exactly verify. The look on the models is still by far more favorable than the last two years.

Do you have the EC and UK precipitation forecasts for ASO since 2016? I would like to compare them to the actual precipitation values for purpose of verification.

[USTropics]

These precipitation plumes are sometimes a good tool to use as an indicator. The last 2-3 seasons it did pretty well indicating overall storm clusterings and tracks.

Not really. TCs are small scale features that cannot accurately be picked up by monthly averaged precipitation anomaly maps. The last two seasons the Euro showed a bone dry Africa and MDR that did not exactly verify. The look on the models is still by far more favorable than the last two years.

Do you have the EC and UK precipitation forecasts for ASO since 2016? I would like to compare them to the actual precipitation values for purpose of verification.

For ECMWF, I'd only go back to 2017 honestly (when SEAS5 became operational). Previous precipitation forecasts with SEAS4 were of even poorer forecast skill (SEAS5 had substantial ocean-atmosphere coupling upgrades and resolution enhancements). Despite this, there has remained a heavy bias against tropical precipitation in the March-May forecasts for JAS (June/July forecasts tend to have higher skill after spring predictability barrier ends for ASO).

Here are the past 3 years of May forecasts of global precipitation for JAS and the verified surface precipitation for the same JAS of each year:

2017 May ECMWF forecast for JAS


2017 observed precipitation for JAS


2018 May ECMWF forecast for JAS


2018 observed precipitation for JAS


2019 May ECMWF forecast for JAS


2019 observed precipitation for JAS


Animation of past 4 May forecasts for JAS (including current May 2020 forecast):

[Shell Mound]

Not really. TCs are small scale features that cannot accurately be picked up by monthly averaged precipitation anomaly maps. The last two seasons the Euro showed a bone dry Africa and MDR that did not exactly verify. The look on the models is still by far more favorable than the last two years.

Do you have the EC and UK precipitation forecasts for ASO since 2016? I would like to compare them to the actual precipitation values for purpose of verification.

For ECMWF, I'd only go back to 2017 honestly (when SEAS5 became operational). Previous precipitation forecasts with SEAS4 were of even poorer forecast skill (SEAS5 had substantial ocean-atmosphere coupling upgrades and resolution enhancements). Despite this, there has remained a heavy bias against tropical precipitation in the March-May forecasts for JAS (June/July forecasts tend to have higher skill after spring predictability barrier ends for ASO).

Here are the past 3 years of May forecasts of global precipitation for JAS and the verified precipitation rates for the same JAS of each year:

2017 May ECMWF forecast for JAS
https://i.imgur.com/6Ci8Agc.png

2017 observed precipitation for JAS
https://i.imgur.com/PeTnnP4.png

2018 May ECMWF forecast for JAS
https://i.imgur.com/8CFZHwO.png

2018 observed precipitation for JAS
https://i.imgur.com/qwFenRE.png

2019 May ECMWF forecast for JAS
https://i.imgur.com/WlPJ0vc.png

2019 observed precipitation for JAS
https://i.imgur.com/sHcVW6c.png

Animation of past 4 May forecasts for JAS (including current May 2020 forecast):
https://i.imgur.com/zXvAs0K.gif

The latter two EC charts for 2018 and 2019, respectively, only extend to JJA, not JAS. Also, the ESRL charts only cover JJA, not JAS.

Wouldn’t ASO be a better time frame to cover, since it encompasses the traditional peak months of the Atlantic hurricane season?

One final point: precipitable water does not measure the amount at the surface, but in the column; precipitation rate does.

[USTropics]

Do you have the EC and UK precipitation forecasts for ASO since 2016? I would like to compare them to the actual precipitation values for purpose of verification.

For ECMWF, I'd only go back to 2017 honestly (when SEAS5 became operational). Previous precipitation forecasts with SEAS4 were of even poorer forecast skill (SEAS5 had substantial ocean-atmosphere coupling upgrades and resolution enhancements). Despite this, there has remained a heavy bias against tropical precipitation in the March-May forecasts for JAS (June/July forecasts tend to have higher skill after spring predictability barrier ends for ASO).

Here are the past 3 years of May forecasts of global precipitation for JAS and the verified precipitation rates for the same JAS of each year:

2017 May ECMWF forecast for JAS
https://i.imgur.com/6Ci8Agc.png

2017 observed precipitation for JAS
https://i.imgur.com/PeTnnP4.png

2018 May ECMWF forecast for JAS
https://i.imgur.com/8CFZHwO.png

2018 observed precipitation for JAS
https://i.imgur.com/qwFenRE.png

2019 May ECMWF forecast for JAS
https://i.imgur.com/WlPJ0vc.png

2019 observed precipitation for JAS
https://i.imgur.com/sHcVW6c.png

Animation of past 4 May forecasts for JAS (including current May 2020 forecast):
https://i.imgur.com/zXvAs0K.gif

The latter two EC charts for 2018 and 2019, respectively, only extend to JJA, not JAS. Also, the ESRL charts only cover JJA, not JAS.

Wouldn’t ASO be a better time frame to cover, since it encompasses the traditional peak months of the Atlantic hurricane season?

One final point: precipitable water does not measure the amount at the surface, but in the column; precipitation rate does.

ECMWF forecasts should all be fixed to JAS (ending month is included in composite data). While ASO is peak for Atlantic hurricane activity, JAS is peak for MDR activity. I'll throw together ASO though in a bit.

[Shell Mound]

For ECMWF, I'd only go back to 2017 honestly (when SEAS5 became operational). Previous precipitation forecasts with SEAS4 were of even poorer forecast skill (SEAS5 had substantial ocean-atmosphere coupling upgrades and resolution enhancements). Despite this, there has remained a heavy bias against tropical precipitation in the March-May forecasts for JAS (June/July forecasts tend to have higher skill after spring predictability barrier ends for ASO).

Here are the past 3 years of May forecasts of global precipitation for JAS and the verified precipitation rates for the same JAS of each year:

2017 May ECMWF forecast for JAS
https://i.imgur.com/6Ci8Agc.png

2017 observed precipitation for JAS
https://i.imgur.com/PeTnnP4.png

2018 May ECMWF forecast for JAS
https://i.imgur.com/8CFZHwO.png

2018 observed precipitation for JAS
https://i.imgur.com/qwFenRE.png

2019 May ECMWF forecast for JAS
https://i.imgur.com/WlPJ0vc.png

2019 observed precipitation for JAS
https://i.imgur.com/sHcVW6c.png

Animation of past 4 May forecasts for JAS (including current May 2020 forecast):
https://i.imgur.com/zXvAs0K.gif

The latter two EC charts for 2018 and 2019, respectively, only extend to JJA, not JAS. Also, the ESRL charts only cover JJA, not JAS.

Wouldn’t ASO be a better time frame to cover, since it encompasses the traditional peak months of the Atlantic hurricane season?

One final point: precipitable water does not measure the amount at the surface, but in the column; precipitation rate does.

ECMWF forecasts should all be fixed to JAS (ending month is included in composite data). While ASO is peak for Atlantic hurricane activity, JAS is peak for MDR activity. I'll throw together ASO though in a bit.

Thank you for clarifying. For the ESRL data, why are you using precipitable water instead of surface precipitation rate (mm/day?) Isn’t the EC using the latter?

[USTropics]

Also since we're using anomalies of precipitable water at the surface, they're completely fine to use as comparisons. Precipitation rate anomaly charts show much less spatial data.

[USTropics]

Also will point out, the NCEP analysis uses a climate period of 1981-2010. The ECMWF is using 1993-2016 I believe. That also makes these comparisons not exact, but overall you can see the ECMWF forecasts in spring tend to have a bias against tropical moisture in the global tropical regions (not just Atlantic, but most prevalent there).

[USTropics]

For UKMET, here are animated graphics that transition from JAS to ASO for May forecasts:

May 2017 forecast

May 2018 forecast

May 2019 forecast

May 2020 forecast


If you want to check out these graphics (and other parameters), you can find them here: https://climate.copernicus.eu/seasonal-forecasts
Composite graphics are created through NOAA's PSL database: https://psl.noaa.gov/cgi-bin/data/getpage.pl
- Monthly/seasonal: https://psl.noaa.gov/cgi-bin/data/compo ... intpage.pl
- Daily: https://psl.noaa.gov/data/composites/day/

[Hammy]

There was a large continuing Central American Gyre in 1998 during much of late May through mid-July that sent several strong waves (that were only kept in check by the waning El Nino) into Texas/Louisiana and I'm getting some similar vibes with this pattern that's expected to take hold.

[CyclonicFury]

Here is how precipitation anomalies actually verified over the Atlantic and Africa from ASO 2017-19 compared to their May ECMWF ASO forecasts:

2017
Precipitation in the MDR east of the Caribbean during ASO 2017 was actually less than average, while in the western Caribbean - where no hurricanes formed - precipitation was above average! Africa was also slightly wetter than normal, which the Euro failed to predict.



2018
Precipitation in the MDR was near average in 2018 overall, and actually above average east of 40W. Precipitation in the Caribbean was generally near to below average. However, the Euro predicted near-average rainfall in the Sahel region of Africa, while much above-average rainfall verified. The ECMWF completely missed the strong West African Monsoon in 2018, which was likely the main reason why the MDR was fairly active despite being much cooler than normal for most of the season.



2019
The MDR was slightly drier than normal in 2019 (and near normal over the eastern MDR), but the Caribbean was much drier than normal. The ECMWF was generally accurate with its precipitation anomaly forecasts for the MDR/Caribbean, but once again, it was way off with precipitation over western Africa (which was much above average, contrary to the ECMWF's prediction of below average).



The ECMWF has generally been too dry in the MDR, but it has been way too dry over western Africa, the source of Atlantic tropical waves. The ECMWF has been underestimating the African Monsoon the past three years, which is likely why its hurricane activity predictions have been too low.

The point is that precipitation anomaly forecasts on models are not meant to be accurate representations of hurricane tracks, but they are an averaged precipitation anomaly over a three-month timeframe. 2017 is a prime example, with below-average precip in the MDR (where three major hurricanes formed) and above-average precip in the western Caribbean (where no hurricanes formed). Just because modeled three-month precipitation anomalies are above average in a region does not mean several hurricanes will form, and just because modeled three-month precipitation anomalies are below average does not mean that no tropical cyclones will form. Tropical cyclones are small-scale features that typically are only in a particular region over a short time, and their signal does not typically show up over a monthly or trimonthly average precipitation anomaly map.

[USTropics]

Here is how precipitation anomalies actually verified over the Atlantic and Africa from ASO 2017-19 compared to their May ECMWF ASO forecasts:

2017
Precipitation in the MDR east of the Caribbean during ASO 2017 was actually less than average, while in the western Caribbean - where no hurricanes formed - precipitation was above average! Africa was also slightly wetter than normal, which the Euro failed to predict.
https://i.imgur.com/Nn37ThZ.png
https://i.imgur.com/kjnMpxG.png

2018
Precipitation in the MDR was near average in 2018 overall, and actually above average east of 40W. Precipitation in the Caribbean was generally near to below average. However, the Euro predicted near-average rainfall in the Sahel region of Africa, while much above-average rainfall verified. The ECMWF completely missed the strong West African Monsoon in 2018, which was likely the main reason why the MDR was fairly active despite being much cooler than normal for most of the season.
https://i.imgur.com/Elqdyhn.png
https://i.imgur.com/ELry9R0.png

2019
The MDR was slightly drier than normal in 2019 (and near normal over the eastern MDR), but the Caribbean was much drier than normal. The ECMWF was generally accurate with its precipitation anomaly forecasts for the MDR/Caribbean, but once again, it was way off with precipitation over western Africa (which was much above average, contrary to the ECMWF's prediction of below average).
https://i.imgur.com/PL1R9U0.png
https://i.imgur.com/ilnrfzl.png

The ECMWF has generally been too dry in the MDR, but it has been way too dry over western Africa, the source of Atlantic tropical waves. The ECMWF has been underestimating the African Monsoon the past three years, which is likely why its hurricane activity predictions have been too low.

The point is that precipitation anomaly forecasts on models are not meant to be accurate representations of hurricane tracks, but they are an averaged precipitation anomaly over a three-month timeframe. 2017 is a prime example, with below-average precip in the MDR (where three major hurricanes formed) and above-average precip in the western Caribbean (where no hurricanes formed). Just because modeled three-month precipitation anomalies are above average in a region does not mean several hurricanes will form, and just because modeled three-month precipitation anomalies are below average does not mean that no tropical cyclones will form. Tropical cyclones are small-scale features that typically are only in a particular region over a short time, and their signal does not typically show up over a monthly or trimonthly average precipitation anomaly map.

Really good analysis. When we're analyzing these precipitation forecasts, we have to understand these are ocean-atmosphere coupling model forecasts. Essentially these anomalies are responses to large scale oscillations in the atmosphere (e.g. ENSO, SOI, NAO, etc.), not a response to tropical systems. I would add that these forecasts can be used to look at how active season patterns were and nonactive patterns. If you're looking at precipitation rate, precipitable water, or whatever parameter you use, what we're really asking is how much moisture is already available in the atmosphere for these tropical systems to work with? Relative humidity is the best parameter, but there are other means to analyze this of course.

TL: DR Use precipitation forecasts to compare to previous active/nonactive seasons, as these are responses to large scale changes in the atmosphere not mesoscale identities like tropical systems. More available moisture in the atmosphere over the MDR (likely due to ENSO forcing and other factors) simply means these systems have a more suitable environment for development (when all other parameters like shear, SSTs, upper divergence, etc. are also favorable).

10 most active seasons in the Atlantic (by ACE) SPR:


Relative humidity:


10 least active seasons in the Atlantic (by ACE) SPR:

[NotSparta]

Here is how precipitation anomalies actually verified over the Atlantic and Africa from ASO 2017-19 compared to their May ECMWF ASO forecasts:

-snip-

Really good analysis. When we're analyzing these precipitation forecasts, we have to understand these are ocean-atmosphere coupling model forecasts. Essentially these anomalies are responses to large scale oscillations in the atmosphere (e.g. ENSO, SOI, NAO, etc.), not a response to tropical systems. I would add that these forecasts can be used to look at how active season patterns were and nonactive patterns. If you're looking at precipitation rate, precipitable water, or whatever parameter you use, what we're really asking is how much moisture is already available in the atmosphere for these tropical systems to work with? Relative humidity is the best parameter, but there are other means to analyze this of course.

TL: DR Use precipitation forecasts to compare to previous active/nonactive seasons, as these are responses to large scale changes in the atmosphere not mesoscale identities like tropical systems. More available moisture in the atmosphere over the MDR (likely due to ENSO forcing and other factors) simply means these systems have a more suitable environment for development (when all other parameters like shear, SSTs, upper divergence, etc. are also favorable).

10 most active seasons in the Atlantic (by ACE) SPR:
https://i.imgur.com/Engf1my.png

Relative humidity:
https://i.imgur.com/S6mKHax.png

10 least active seasons in the Atlantic (by ACE) SPR:

https://i.imgur.com/dJlFWO2.png

Interesting to note that the least active seasons tend to have below average precip in the far eastern Pacific. I believe active seasons actually have above average precip there as it enhances the ATL ITCZ and also signals that most TC activity is in the eastern part of the basin signalling unfavorable conditions in the EPAC

[chaser1]

I've copy/pasted this dialogue from the ENSO thread given it's relevance to evolving 2020 Indicators

Another post from Ventrice.

So Ventrice went on to talk about an "Atlantic La Nina" also forming. I have to admit that I haven't read much about that phenomenon but he suggested it would strengthen or enhance the WAM. Not sure what that would signal for the MDR or Atlantic Basin on a whole. He did not elaborate.

http://twitter.com/MJVentrice/status/1264525876840083457?s=20
http://twitter.com/MJVentrice/status/1264526432212000771?s=20

The general implication is that the "WAM" (West African Monsoon) could be on overdrive, suggesting increased tropical wave activity. What i'm not entirely sure about would be whether this would or would not suggest a more enhanced ITCZ throughout the MDR. I tend to think that an enhanced WAM would certainly aid toward the expectation of a higher aggregate number of tropical cyclones within the basin but such a robust low level flow might inhibit much of the development until such waves reach a longitude where either the lower flow might slacken a bit (or where vertical shear may be more relaxed). Overall I think that a strong WAM portends of a good deal more activity and perhaps storm tracks through the Caribbean then displayed during recent years.

[Shell Mound]

Based on current NAO forecasts (both ECMWF- and GFS-based), the above-average trade winds over the MDR don’t seem to slacken for at least the next week or so. This means that we are unlikely to see significant short-term change (warming) in the MDR over the next ten days. The major problem is that the equatorial low-level westerlies just can’t seem to lift northward, owing to stubborn mid-level divergence (col-type pattern), which results in low-level sinking, implying above-average MSLP and associated easterlies over the MDR. So we have a weird pattern, a state of bipolarity, in which the equatorial and subtropical Atlantic is in a +AMO, but the MDR, Caribbean, and far North Atlantic (near Greenland) are in more of a -AMO, resulting in a blend of EOF1/2, as others have mentioned previously. It will be interesting to see how the strong African monsoon will counteract the effect of the -AMO-type influence(s) and interact with the MJO, another factor that could easily make the difference between hyperactivity and something in line with a “normal” active season. Overall, as in the past several seasons, there seems to be higher-than-normal uncertainty well beyond the end of the SPB.

[StruThiO]

 https://twitter.com/BenNollWeather/status/1264938020945358851

[StruThiO]

Interesting.

 https://twitter.com/EricBlake12/status/1264954409747918848