Why Did The Tropical Atlantic Break Dry Air Record For August And September?

[bluewave]

August and September saw the driest air over the Tropical Atlantic on record for those two months.

While dry air has been increasing especially since the 2006, this was the first hurricane season

so strongly impacted by the pattern. The dry air slowly began developing over South America from

around 1995-2005 and spread out toward the Tropical Atlantic after 2005. The 8 year Atlantic ACE

has dropped from 1308 during 1998-2005 to 799 from 2006-2013. It's interesting that South America

also experienced an unusual amount of dry air during the 1950's and 1960's which extended to

the Eastern Pacific. 

 

 

 

 

 

 

 

 

 

 

 

 

[snowmanwx]

Interesting.  If I may interject a little geopolitical perspective, consider the Sahel region of Africa in this progression of images.  After several decades of unusual wetness, a drought commenced in 1968, leading to tremendous human suffering from famine and severe population dislocations.  Given the seemingly repeating cycle in these graphics, the Sahel region may return to severe drought over the next few decades.  If this return to drought occurs, then a heavily Islamic population of hundreds of millions would get desperate, but international boundaries inhibit the movement of populations southward into wetter climates.  The population of most countries in the region has tripled or even quadrupled since 1968, and millions of already often malnourished poor Africans dwell in immense urban slums that essentially did not exist a half-century ago.   The progression of drought to this region likely would lead to dramatic increases in local food prices that could trigger major geopolitical repercussions as enormous displaced or desperate population seek sustenance and refuge abroad. 

 

All that said, the ongoing positive phase of the Atlantic multi-decadal oscillation ultimately must come to a close.  Still, its historical periodicity suggests that the close might should occur a decade or more into the future.  Perhaps then, this year represents a pause in a series of hyper-active Atlantic seasons.  We shall see which course future years will take. 

[WXheights]

With prevailing low latitude easterly waves, would have thought the Sahara would have been the source region. The correlation to South America (Amazon) opens new questions.

[icebreaker5221]

With prevailing low latitude easterly waves, would have thought the Sahara would have been the source region. The correlation to South America (Amazon) opens new questions.

 

Or perhaps the source region was from above...

 

[OKpowdah]

Time series of 500mb RH averaged across the MDR for Aug-Sep

 

[jordanwx]

On another site (wunderground) I outlined some reasons for the dry air, lack of vertical instability and downward motion over our basin:

 

In the global tropics, we don't have an El Nino, the Indian ocean and SE Asia region (monsoon) is not anomalous in a bad way, the African monsoon is very healthy, and SSTAs in the tropical Atlantic are above average. All the things we normally look for as problems are not problems. This means we have to look in places we wouldn't normally look. Here are a few problems I have found in plotting at PSD for Jul-Sep:

 

1. Large zonal band of upward motion over the Pacific just off the equator. This upward motion band is not indicative of an El Nino, but such a large band of upward motion right next door to our basin can't be good.

 

 

 

 

2. Broad Hadley cell in the North Atlantic (meridional overturning). A lack of convergence coming into the MDR from the subtropical high in the NATL at the low levels. The lowest levels of the atmosphere (from 850-sfc) experienced lower (anomalous) pressures in the subtropics than the deep tropics, robbing the tropics of their convergence. In the upper levels the problem is also evident by a higher height anomalies over the deep tropics and lower height anomalies over the subtropics. The subtropical SSTA profile and PDO are responsible for this.

 

 

3. Subtropical ridging in the South Atlantic. Image. Ridging persisted in the central subtropical South Atlantic all season long. Strong subsidence was created on the backside of the ridge (positive omega off Africa below) advecting large amounts of anomalously dry air directly into the tropical Atlantic (second image).

 

Omega Anomaly

 

 

 

Vector Wind Anomaly

 

 

4. Global Warming/Climate Change. 500mb RH has been on the downtrend for the last several decades as OKpowdah plotted above.

[jordanwx]

Time series of 500mb RH averaged across the MDR for Aug-Sep

 

climindex.72.198.25.44.295.11.25.12.png

Wow, that is very telling. Looks like gw/cc should be added to the list of possible causes.

[am19psu]

Nice write up, jordan.

[bluewave]

Time series of 500mb RH averaged across the MDR for Aug-Sep

 

climindex.72.198.25.44.295.11.25.12.png

 

You can also see the increasing 500mb RH on the EPAC side against the MDR decrease over the same period of time. There is also a rising moisture pattern peaking over Northern South America around 1990 before falling again.

 

 

 

[icebreaker5221]

Obviously the prevailing wind at 500 mb over the MDR was still east to west in 2013 (as per climo) and disturbances came off of Africa. However, as the figures originally posted by bluewave showed, the air over Africa was actually more moist than normal. Therefore, it is unlikely Africa alone was a source for such anomalously dry air over the MDR. On the other hand, air over S America was anomalously dry. However, it was also not likely a dominant source of dry air (although it may have been a contributor at times), because (1) the mean flow was not in general from S America towards the MDR and (2) the region is climatologically very moist, so even anomalously dry air coming from S America is not that dry at all compared to SAL air or mid-latitude air.

The more likely culprit is east to west moving tropical waves that encounter large scale subsidence, retaining mixing ratio but warming dry adiabatically such that resulting saturation fraction is lowered.

500 mb mean wind vector, Aug-Sep 2013:

[Drz1111]

Agreed with the below.  I'm no met, but I suspect humidity (as it generally is in the tropics) is a proxy for subsidence.    I'm not surprised that you'd observe lots of sinking air in a season with few cyclones, and query whether you're just observing a tautology - i.e., rather than "sinking air caused fewer cyclones", this seasonal average appears to have more sinking air because there were so few cyclones (and attendant rising air) passing by during the applicable period.

 

 

 

Obviously the prevailing wind at 500 mb over the MDR was still east to west in 2013 (as per climo) and disturbances came off of Africa.  However, as the figures originally posted by bluewave showed, the air over Africa was actually more moist than normal.  Therefore, it is unlikely Africa alone was a source for such anomalously dry air over the MDR.  On the other hand, air over S America was anomalously dry.  However, it was also not likely a dominant source of dry air (although it may have been a contributor at times), because (1) the mean flow was not in general from S America towards the MDR and (2) the region is climatologically very moist, so even anomalously dry air coming from S America is not that dry at all compared to SAL air or mid-latitude air. 

 

The more likely culprit is east to west moving tropical waves that encounter large scale subsidence, retaining mixing ratio but warming dry adiabatically such that resulting saturation fraction is lowered. 

 

500 mb mean wind vector, Aug-Sep 2013:

iHWgddACQg.png

[icebreaker5221]

Agreed with the below.  I'm no met, but I suspect humidity (as it generally is in the tropics) is a proxy for subsidence.    I'm not surprised that you'd observe lots of sinking air in a season with few cyclones, and query whether you're just observing a tautology - i.e., rather than "sinking air caused fewer cyclones", this seasonal average appears to have more sinking air because there were so few cyclones (and attendant rising air) passing by during the applicable period.

 

Well, first let me comment that seasons with lower ambient MSLP across the MDR have been correlated with higher TC activity.  Yes, when you do a retrospective season analysis the fact that there were more TCs lowers the overall MSLP for the season, but you'll nonetheless generally get lower than average MSLP even if you exclude the times and locations of TCs (I believe Klotzbach and maybe a few others have run the numbers).

 

Although I have not run the numbers, I strongly suspect omega would be the same.  Seasons with more ambient/environmental upward motion are more favorable for TCs, and the existence of those TCs subsequently decreases the seasonal-mean omega over the region beyond what it otherwise might be. 

[jordanwx]

The more likely culprit is east to west moving tropical waves that encounter large scale subsidence, retaining mixing ratio but warming dry adiabatically such that resulting saturation fraction is lowered.

500 mb mean wind vector, Aug-Sep 2013:

iHWgddACQg.png

Yep, but the question remains, "what is causing that?" See my comments above for an answer to that.

[skierinvermont]

I'm not sure long-term trends in upper level RH from NCEP are considered reliable..

 

I know that globally RH is believed to have stayed constant meaning that a regional decline in the Atlantic is probably either caused by something other than global warming, or it is a regional result of global warming. 

[jordanwx]

I'm not sure long-term trends in upper level RH from NCEP are considered reliable..

 

I know that globally RH is believed to have stayed constant meaning that a regional decline in the Atlantic is probably either caused by something other than global warming, or it is a regional result of global warming. 

Not sure what level you are referring to, but above 700mb global RH has dropped. This is especially true if we look at just the tropics (what we care about).

 

[Mike.Ventrice]

Not sure what level you are referring to, but above 700mb global RH has dropped. This is especially true if we look at just the tropics (what we care about).

 

 

There is a clear downward trend, but I'm always skeptical of data prior to 1978. It seems while there is a still a downward trend post 1978, it's no where as strong as the negative slope when using pre-satellite era data. 

[Tropopause_Fold]

On another site (wunderground) I outlined some reasons for the dry air, lack of vertical instability and downward motion over our basin:

 

In the global tropics, we don't have an El Nino, the Indian ocean and SE Asia region (monsoon) is not anomalous in a bad way, the African monsoon is very healthy, and SSTAs in the tropical Atlantic are above average. All the things we normally look for as problems are not problems. This means we have to look in places we wouldn't normally look. Here are a few problems I have found in plotting at PSD for Jul-Sep:

 

1. Large zonal band of upward motion over the Pacific just off the equator. This upward motion band is not indicative of an El Nino, but such a large band of upward motion right next door to our basin can't be good.

 

 

 

 

2. Broad Hadley cell in the North Atlantic (meridional overturning). A lack of convergence coming into the MDR from the subtropical high in the NATL at the low levels. The lowest levels of the atmosphere (from 850-sfc) experienced lower (anomalous) pressures in the subtropics than the deep tropics, robbing the tropics of their convergence. In the upper levels the problem is also evident by a higher height anomalies over the deep tropics and lower height anomalies over the subtropics. The subtropical SSTA profile and PDO are responsible for this.

 

 

3. Subtropical ridging in the South Atlantic. Image. Ridging persisted in the central subtropical South Atlantic all season long. Strong subsidence was created on the backside of the ridge (positive omega off Africa below) advecting large amounts of anomalously dry air directly into the tropical Atlantic (second image).

 

Omega Anomaly

 

 

 

Vector Wind Anomaly

 

 

4. Global Warming/Climate Change. 500mb RH has been on the downtrend for the last several decades as OKpowdah plotted above.

this is a good write-up.

 

few thoughts/questions: how would these issues you outline above translate to the western part of the basin and higher in latitude? i feel like there has to be more to the puzzle than this as even periods that appeared favorable for W Carib/W Atlantic/BOC/GOM development struggled mightily to produce. would the broader hadley cell result in a disturbed walker circulation? 

[Sunny and Warm]

Not sure what level you are referring to, but above 700mb global RH has dropped. This is especially true if we look at just the tropics (what we care about).

 

How do you account for the nadir in in 2004-05 and the large number of TC's during those years?  It appears that if you superimposed ENSO signal over that graph, both TC activity and RH may correlate well. 

[bluewave]

How do you account for the nadir in in 2004-05 and the large number of TC's during those years?  It appears that if you superimposed ENSO signal over that graph, both TC activity and RH may correlate well. 

 

The dry air in 2005 wasn't able to extend far enough north to have an impact on that hyperactive season.

Notice how the dry air that year remained closer to South America.

 

 

[HM]

Obviously the prevailing wind at 500 mb over the MDR was still east to west in 2013 (as per climo) and disturbances came off of Africa. However, as the figures originally posted by bluewave showed, the air over Africa was actually more moist than normal. Therefore, it is unlikely Africa alone was a source for such anomalously dry air over the MDR. On the other hand, air over S America was anomalously dry. However, it was also not likely a dominant source of dry air (although it may have been a contributor at times), because (1) the mean flow was not in general from S America towards the MDR and (2) the region is climatologically very moist, so even anomalously dry air coming from S America is not that dry at all compared to SAL air or mid-latitude air.

The more likely culprit is east to west moving tropical waves that encounter large scale subsidence, retaining mixing ratio but warming dry adiabatically such that resulting saturation fraction is lowered.

500 mb mean wind vector, Aug-Sep 2013:

iHWgddACQg.png

 

I'm not really understanding your use of the mean wind here. First of all, there is a meridional component even in your mean graphic which is cross-equatorial. By using anomaly instead, you will see that clearly. There is most certainly a SH influence here, since the subsidence zone sits awfully close to the equator, on average. 

 

At the surface, the two "drier" sources are the Bermuda High and the Saharan Desert. At 500, the dry sources are from the Iranian Plateau to the eastern Atlantic and the SH Atlantic between 0-30 S. Clearly, these sources worked together to destroy the typical equivalent potential temperature values over the Atlantic Basin. There is actually a flip that occurs between the early 2000s and this decade in anomalous equatorial circulation near Africa from cyclonic to anticyclonic. Global circulation has been significantly altered in the Tropical Atlantic-African Sector.

[icebreaker5221]

I'm not really understanding your use of the mean wind here. First of all, there is a meridional component even in your mean graphic which is cross-equatorial. By using anomaly instead, you will see that clearly. There is most certainly a SH influence here, since the subsidence zone sits awfully close to the equator, on average. 

 

At the surface, the two "drier" sources are the Bermuda High and the Saharan Desert. At 500, the dry sources are from the Iranian Plateau to the eastern Atlantic and the SH Atlantic between 0-30 S. Clearly, these sources worked together to destroy the typical equivalent potential temperature values over the Atlantic Basin. There is actually a flip that occurs between the early 2000s and this decade in anomalous equatorial circulation near Africa from cyclonic to anticyclonic. Global circulation has been significantly altered in the Tropical Atlantic-African Sector.

 

If I recall correctly, the original quote that I was refuting (which has subsequently been deleted) was making the argument that the dry air was actually coming from northern South America and/or west Africa.  I can accept your idea that dry air from the south Atlantic from 10-30 W advecting anomalously far north was a significant contributing factor.  Not sure I see as clear a connection from the Iranian Plateau in 2013, at least with respect to the deep tropical Atlantic (say, south of 15 N), but cross-equatorial flow from the south Atlantic was clearly an issue. 

 

That said, the reason it's so dry along the equator from 10-30 W is that, as you said, it is a climatologically-favored region of subsidence.  So with anomalous southerlies in this region, and the entire pattern (ITCZ, equatorial subsidence, hadley cell, etc) shifted slightly north, is it simply dry air blowing northward or is it that the subsidence is displaced further northward?  Maybe a little bit of both?  That would be my guess. 

 

Edit to add climatological RH at 500 mb for reference:

[uncle W]

the 50's and 60's had some years with little tropical action...from 1956 to 1965 there were five years with less than 10 storms...NYC's driest 10 year period was from 1956-1965...

[jordanwx]

How do you account for the nadir in in 2004-05 and the large number of TC's during those years?  It appears that if you superimposed ENSO signal over that graph, both TC activity and RH may correlate well. 

What do you mean?

 

I'm not correlating global RH with Atlantic activity. You stated that global RH has been steady, I'm saying that in the mid-levels and above it has dropped globally, as well as in the global tropics (image I posted). So the downward trend in mid-level Atlantic RH may be a result of global warming among other things.

[jordanwx]

If I recall correctly, the original quote that I was refuting (which has subsequently been deleted) was making the argument that the dry air was actually coming from northern South America and/or west Africa.  I can accept your idea that dry air from the south Atlantic from 10-30 W advecting anomalously far north was a significant contributing factor.  Not sure I see as clear a connection from the Iranian Plateau in 2013, at least with respect to the deep tropical Atlantic (say, south of 15 N), but cross-equatorial flow from the south Atlantic was clearly an issue. 

 

That said, the reason it's so dry along the equator from 10-30 W is that, as you said, it is a climatologically-favored region of subsidence.  So with anomalous southerlies in this region, and the entire pattern (ITCZ, equatorial subsidence, hadley cell, etc) shifted slightly north, is it simply dry air blowing northward or is it that the subsidence is displaced further northward?  Maybe a little bit of both?  That would be my guess. 

 

Edit to add climatological RH at 500 mb for reference:

iIYD98TFGz.png

A blogger on Wunderground was trying to make this argument (that dry air was being advected off SA). I told him this was not possible by showing him the same chart (mean flow). Pretty clear to me dry air isn't coming from South America.

[jordanwx]

this is a good write-up.

 

few thoughts/questions: how would these issues you outline above translate to the western part of the basin and higher in latitude? i feel like there has to be more to the puzzle than this as even periods that appeared favorable for W Carib/W Atlantic/BOC/GOM development struggled mightily to produce. would the broader hadley cell result in a disturbed walker circulation? 

Strong band of upward motion (blue colors) exists in the EPAC and another large area in the western Atlantic basin north of 20N (off the East coast). Focusing the upward motion north of the Caribbean and in the epac will force downward motion (red colors) to exist over the Caribbean. Just conservation of mass.

 

[Sunny and Warm]

What do you mean?

 

I'm not correlating global RH with Atlantic activity. You stated that global RH has been steady, I'm saying that in the mid-levels and above it has dropped globally, as well as in the global tropics (image I posted). So the downward trend in mid-level Atlantic RH may be a result of global warming among other things.

where did I say that?

 

As for causation, I would hesitate to lay blame on GW (I know you are not) until we have good data over a couple multi-decadel cycles.  That could take another 40-60 years to bear fruit.

[jordanwx]

where did I say that?

 

As for causation, I would hesitate to lay blame on GW (I know you are not) until we have good data over a couple multi-decadel cycles.  That could take another 40-60 years to bear fruit.

Woops, someone else did. I got confused cause you responded to my post which was not directed at you.

[HM]

 

If I recall correctly, the original quote that I was refuting (which has subsequently been deleted) was making the argument that the dry air was actually coming from northern South America and/or west Africa.  I can accept your idea that dry air from the south Atlantic from 10-30 W advecting anomalously far north was a significant contributing factor.  Not sure I see as clear a connection from the Iranian Plateau in 2013, at least with respect to the deep tropical Atlantic (say, south of 15 N), but cross-equatorial flow from the south Atlantic was clearly an issue. 

 

That said, the reason it's so dry along the equator from 10-30 W is that, as you said, it is a climatologically-favored region of subsidence.  So with anomalous southerlies in this region, and the entire pattern (ITCZ, equatorial subsidence, hadley cell, etc) shifted slightly north, is it simply dry air blowing northward or is it that the subsidence is displaced further northward?  Maybe a little bit of both?  That would be my guess.

Oh okay...unfortunately, summertime water vapor pathways in the vertical can throw a wrench into the whole thing as far as advection goes. While I agree that the dry air from South America is not an answer at all, we cannot say with 100% confidence how water vapor flux behaved without a deeper analysis. There are very complicated vertical circulations in and out of the Tropics that can lead to strange pathways that do not quite adhere to mean wind.

I mentioned the Iranian area because I was just talking straight climatology for nearby sources of low and mid level dry air. It is was large source from the eastern Atlantic to the Iranian Plateau.

As for z-circulations, this looks like a combination of Walker and Hadley Cell dynamics. A quick look at the mean h2 velocity potential at 5-5°S/N and 10-20°N shows Walker behavior similar to an El Niño, despite the lack of one. Overall, outside of a few Kelvin Waves, the period from June through early October was characterized by mean convergence throughout the Atlantic. Throw in the anomalous equatorial wave propagation from the Mid Latitudes and SH subsidence shift equatorward (related to the -AAO?) and suddenly things became quite hostile over the Basin.

My mistake in the spring was thinking too linearly. The Indian Ocean did in fact cool some from the prior seasons and did affect circulation. But as it turns out, it seemed to help in the -AAO--equatorward high in the S. Atlantic, which played a major role here in the weak season. Whoops!

[bluewave]

A blogger on Wunderground was trying to make this argument (that dry air was being advected off SA). I told him this was not possible by showing him the same chart (mean flow). Pretty clear to me dry air isn't coming from South America.

 

Where the dry and more stable air  came from may be less important than the fact that both areas shared the same region of subsidence. This was just a more extreme version of the pattern that started to become more obvious

after the 2005 season. But it took until around 2010 for it to show up in increasing Tropical Atlantic and Caribbean

stability readings. 

[jordanwx]

Where the dry and more stable air  came from may be less important than the fact that both areas shared the same region of subsidence. This was just a more extreme version of the pattern that started to become more obvious

after the 2005 season. But it took until around 2010 for it to show up in increasing Tropical Atlantic and Caribbean

stability readings. 

Yes, agreed. Think general subsidence is just happening over SA/MDR. Dry air or subsidence is not coming from South America, however. In fact, drought in South America is usually a sign of an active season.