Monday, November 13, 2017

An updated look at the pattern for the 2017-2018 winter outlook.

I've decided to make a update to the winter outlook I posted back on the 11th of October. If you've followed either of my two Facebook pages or on twitter, then you know, I've been saying for 5 mouths, I've been saying winter 2017-2018 was going to be cooler and have more winter precipitation than our last two winters, and this update will stay that course.

A little about my approach to long range forecasting:

Back when I was saying all of this, The Global Forecast System (GFS), European Computer Forecast Model (ECMWF), Canadian Model (CMC), Navy Global Environmental Model ( NAVGEM), and the Japan Meteorological Agency ( JMA), and their associated ensemble model products, were showing the exact opposite. But like all models, these dynamical models have their issues and bias.  Making a forecast especially a long range seasonal forecast, Involves a lot more than just looking at models.  Many of the tricks and tools I use aren't even taught at any met school anymore, I was blessed to have a few old school profs, along with people showing me things when I did field work, that enable me to know some of these older ways of doing things.  So I take these tricks and old time rules, and apply them to the modern way of doing meteorology.  I think the two make a nice blend. Model interpretation and knowing how to blend models based on their individual weaknesses, an ample understanding of teleconnection processes, pattern recognition, doing the math yourself (not relying on the computer to do the dynamic meteorology calculations for you) (I also use a lot of advanced statistics), A good understanding of the atmosphere,  and some understanding of climatology, are all essential to making accurate forecast .  If forecasting was as easy as looking at some computer models, you wouldn't need me. The only reason I say all of this is: far too many meteorologist, weather enthusiasts, and weather geeks rely on weather models to determine what the weather will do. IMO this is a poor forecasting strategy, that more often than not leads to failure.  I pride myself in giving you the best real analysis you will find anywhere.  There might be paid weather outlets that provide the same service, but I have yet to see a free site that provides a superior product than I do.  

Anyway on to the updated outlook:

No two winters are alike!



El Nino-Southern Oscillation (ENSO):


Like all teleconnections, the ENSO has a positive , neutral, and a negative phase.  The positive phase is called El Nino and the negative phase is called La Nina.  The neutral phase is called La Nada.  La Nada winters are very cold in the Great Lakes, Northeast, into the Middle Atlantic.

Last winter was one that featured above average snow for much of the Northeast. ; while parts of the Mid Atlantic along the I-95 saw much less than average snowfall.  At the same time last winter was one of above average temperatures, in spite of a cold March the winter was well above average in the temperature department.

Right now, we're in a weak La Nina. I've been saying for months that a weak La Nina was likely. La Nina winters tend to favor a very cold winter for the Upper Midwest and Northern Plains, While the Southeast tends to see very warm temperatures. The Middle Atlantic sees more in the way of average temperatures.  As is typical of La Nina, we've already seen a lot of high pressure in southeast Canada. Many times this setup can result in end of November and December ice storms in the Northeast. Last year was a weak La Nina as well. But this year there are a few key differences in the Pacific.  


Quasi-Biennial Oscillation (QBO):


For winter 2016-2017 we started heading for an East based QBO. But, that faded and we ended up with a west based QBO. This winter will have a QBO is that is east based. This will have a  big ramification on this winter season.  The state of the QBO has an subtle difference on La Nina.  When we have an west based QBO,  We have a trough near the Aleutians leading to a cold Alaska and western Canada; also there is a lack of high latitude blocking near Greenland during a west based QBO.  But with an east based QBO we normally see high latitude blocking. Along with a ridge out west. This warms up Alaska and displaces the cold south and east.  The cold rides the jet stream down into the East CONUS. Because of the High latitude blocking the southeast ridge sees more suppression than during a west based QBO.  

Arctic Oscillation (AO) and the North Atlantic Oscillation (NAO):


Both of these teleconnections are kissing cousins .  When we have an east based QBO our winter season is much more prone to having a negative AO and negative NAO. This is the reason for all the high latitude high pressure over Greenland, leading to the upstream blocking.

Pacific  Decadal Oscillation(PDO):


The PDO is looking to be slightly negative to neutral for winter 2017-2018.  When the PDO is positive (warm phase) it can weaken a La Nina. When the PDO is negative (cool phase) it can strengthen a La Nina. 

The warm blob:

If you remember, back in 2013 a lot was beings said about the warm SST south of the Aleutian Islands. They dubbed this the "Warm Blob"  When this area of the Pacific is warm, the East Coast and Mid Atlantic have a much better chance of seeing a colder winter on average.

Warm SST in the Western Atlantic:

The warm SST's off the East Coast will help keep the Mid Atlantic and Southern New England a little warmer. But at the same time, if we do happen to see a Nor'easter form off the Southeast Coast, the warm water would vastly increase snowfall amounts for the big cities.


East Pacific Oscillation (EPO):


We're also having an negative  EPO.  When the EPO is negative this also implies a upper level ridge just off the West Coast and a trough in the east.  A positive EPO is the exact opposite. 

West Pacific Oscillation (WPO):

The WPO is strongly negative


And as I said in my winter outlook released in early October, years that have low solar activity, tend to be cooler during the winter.  Last year was also a low solar year, but the west based QBO worked against it.  For winter 2017-2018 that won't be the case.

An old weather observer who recorded weather events not only for his lifetime, but those of his Father and Grandfather as well, bequeathed his journals to me 12 years ago.  The journals are filled with all kinds of wisdom and observations, that have helped me immeasurably.  One thing they noted was that end of October and November  weather can be a good gauge as to what the winter will be.  If we have a cold October and warm November, on average we see more in the way of mild temperatures from December through March.  But when we have a warm October and a cold November, we often end up with the rest of winter being cooler and snowy.   

The Great Lakes are warm for this time of year.  That could very well play a role in early season lake effect snow events off of Lake Erie and Lake Ontario. In fact we could see a big lake effect event or two before Christmas.  

Looking at years like 1950, 1995,  2005, 2007,2008,2010, 2014 stack up well to what appears to be going on.

These winters seem to be a good match for what we can expect for this winter.

1950/1951,1960,1961,1968/1969, 1984/1985, 1995/1996, 1998/1999, 2005/2006 ,2009/2010, 2014/2015

Accumulated Cyclone Energy (ACE):

The ACE index deals with wind energy. It's used to try determine how active and distractive a particular tropical storm as well as entire hurricane season will be.  

Meteorologist  Joe Bastard and a few other meteorologist , are trying to correlate high ACE years with colder weather from November into March.  What I've seen of all this is interesting. Joe Bastardi of Wxbell, shows that because 2017 was a high ACE year, this coming winter will be cold. I don't know enough about using the ACE index to forecast cold winters. But like I said, the theory is interesting. 

What all of this means:


The next couple of weeks will be a fight between the cold building in the north and warmth in the south.  negative NAO and the positive EPO.  I do still think the EPO will turn negative. The WPO is just too negative for it not to.  


When you input the analog years into the models,  you end up with a similar look to what I show in my outlook.  Years that had two or more teleconnection similarities  November into March tended to end up cooler than average in the Northeast.  The idea of a blend between the current weak La Nina and a La Nada, also supports the idea that this coming winter will be cooler than many think.  When you put an EPO and the warm blob into the equation the trend is even colder. 

While La Nina winters tend to be overall mild for much of the Northeast. The Idea of an EPO and the warm blob will have a big influence on just how mild we will get.  So this winter will end up being a cooler La Nina.  As I've been saying quite frequently on my Facebook page, there is a lot of cold air building  across all of Canada. Because of the better chance for a negative NAO and negative AO, we could end up with a few polar vortex events. So even with a La Nina there is a chance for a few really cold arctic outbreaks.

 Winter will get an early start.  Most of the teleconnections are very strongly suggesting November and December will be cool.  So think most of us will see several chances for snow over the next few weeks heading into Christmas The global models especially the GEFS are having a lot of issues with the current pattern.  This all has to do with their individual biases and feedback issues dealing with the warmth in other areas of the planet. The cold in December should go into at least the first half of January. During January we could see a thaw that could last ten days or more. But the cold should come back into the pattern for February into March.  More or less a bookend winter.   

I expect Pennsylvania, New York State, and New England see the snowpack build. But La Nina winters are precarious, with swings in temperature. But overall I think we will do alright in the snow department.  La Nina winters favor a northern storm track into the Great Lakes and Northeast.  During La Nina we tend to see more clipper type storms then during none La Nina winters.  January to March tend to see quite a few snow chances  in the Upper Midwest and Northern Great Lakes, it also can get quite snowy across  Western Pennsylvania, much of New York State, Vermont, New Hampshire, and Maine. At the same time southern eastern Pennsylvania, New Jersey into the Northern Mid Atlantic  are a little less snowy.

With the active northern stream and the occasional surges north of the Southeast ridge,  winter 2017-2018 will see some ice storms, this will be especially true for the Mid Atlantic into Pennsylvania.  

Because of the active northern storm track, the Middle Atlantic won't see many coastal storms, all or most of their snow will come from clippers. The clippers will deposit snow in the Tennessee and Ohio Valley into New York State and parts of New England. So for Virginia, Delaware, Maryland, and southeast Pennsylvania the snow amount will mostly depend on clipper type storms. But we should still see a relaxing in the pattern from time to time, which would allow the southern stream to get involved in the pattern. So I can't rule out three or four Lows developing near the Gulf and then moving toward the Mid Atlantic and New England.  During a La Nina year the coastal typically form off the Delmarva or south of New England. When there is a strong to moderate La Nina, the Middle Atlantic sees warm temperatures and a lot less snow. However when we have a weak La Nina, snowfall seems to be closer to average.

But a Side note, often when we see a big storm in October, we tend to see at least one bigger storm during the winter.  Something to think about.  All it takes is one big storm to cause the Middle Atlantic to see above average snowfall.  The pattern is less supportive of this idea; but historically October storms can be a precursor to a stormy Mid Atlantic.    

For Pennsylvania west of I -81, much of New York State and New England will end up  overall with above average snowfall and below average temperatures.  most of eastern Pennsylvania east of I-81, temperature will be warmer than western Pennsylvania, but still below average overall.  Southeast Pennsylvania into the Mid Atlantic, will have a greater likelihood of seeing  some issues with the Southeast ridge; So there could end up to be average to slightly above average with overall temperatures for December into February.   I still think overall January will be the coldest month this winter. End of January into February we could see some temperature swings, as the cold pattern relaxes. Then for The rest of February and March we will once again end up below average with temperatures.

This winter will be colder than last winter here in the Northeast. Snowfall will be above average. 

The Middle Atlantic and the I-95 corridor will end up around average in temperatures  Snowfall  should be average to below average for snowfall.



As you can see there isn't a lot of difference between this outlook and my earlier outlook.
Remember a seasonal outlook isn't so much a forecast, as it is a broad brush to help guide us in what should happen.  I can't give you specifics as to what will happen or when things will occur. But generally this is what I think our winter of 2017-2018 will be like.

Wednesday, October 11, 2017

My Winter 2017 2018 outlook

Well it's getting close to that time of year, when thoughts start to shift toward winter, and what might it be like.
One question I've seen a lot is, because our Summer was cool and wet does that mean this winter will be cold and wet?  My answer is sure, there is a chance. But then there is a chance it won't.  For several years, I've looked at the analog data hard; I've never found any direct correlation between a cool summer and a cool winter.

This year I want to start out listing my analog winter seasons. I used these to try an come up with a reasonable idea on how this winter should behave.  

Analog years:

I looked at the years that had a weak La Nina along with a negative QBO.... 1950, 1954, 1956, 1962, 1964,1967, 1970, 1974, 1983, 1988, 1995, 2000, 2005, 2007, and 2011

Analog years ended up being 1952/1953, 1953/1954, 1958/1959,1960/1961, 1962/1963 1964/1965, 1967,1968, 1969/1970, 1970/1971, 1975/1975, 1977/1978, 1979/1980, 1981/1982,1995/1996, 2000/2001 2004/2005, 2006/2007, 2007/2008, 2014/2015.  I gave the seasons in red the most weight, with  1974/1975, 1981/1982 and 20006/2007 being double weighted.

The thing that directly impacts our weather is the jet stream. To refresh Y'all, the jet stream is a narrow river of fast moving air that resides above us in the upper atmosphere. There are many jet streams, but the two that effect the Northeast the most are the sub tropical jet and the polar jet.  The polar jet sits around the 500mb level (25,000-30,000 feet), while the subtropical jet is a little higher, around the 300mb to 115mb level (30,000 to 50,000 feet ) During the Summer the polar jet is typically found well to the north in northern Alaska and northern Canada. While the subtropical jet pushes south to near the Gulf Coast.

The wavering of these two jet streams is what causes the day to day variations in our weather. When the polar jet drifts south, we see cooler weather. When the subtropical jet moves north we see warmer weather. 

But our weather is also effected by other teleconnection flows. the El Nino and the Southern Oscillation, the North Atlantic Oscillation. the Quassi-Biennial Oscillation, The Pacific Decadal Oscillation, and others. We also are effected by things like sea surface temperatures, snow and ice extent, and even sun spots. All of these things play a role in how cold and snowy our winter will be.

Sea Surface Temperatures and the El Nino and the Southern Oscillation (ENSO):

There are four regions in the tropical pacific that we track, El Nino region 1+2, 3, 3.4, and 4.

I think Winter 2017-2018 will be more of a negative ENSO or a weak La Nina. The La Nina could get close to moderate, but I rather doubt it. The SST in Nino regions 1,2 3, 3.4 were dropping off fairly quickly. Then there was a warming trend. The last couple of weeks have seen the Nino region start to cool off again. They are currently below normal and are approaching La Nina levels. Because of the up and down cooling most atmospheric indicators have remained ENSO neutral. Winter 2016-2017 was a mix of El Nino and La Nina.

The type, timing and strength of the La Nina will be critical. an east or central based is more favorable for a cold east of the Rockies, compared to a west based event.
Sea Surface temperatures in the northeast Pacific and off the Eastern Seaboard of the U.S.

Currently, sea surface temperatures (SST) in the eastern tropical Pacific are slightly below average. The mid eastern Pacific has seen the biggest departure from average, right now that's about one degree Celsius.  The coolness does extend down a few hundred meters below the surface. We should see the east - central Nino regions will continue to cool . Last winter, regions 1 and 2 got warm, I don't see them getting that warm this winter.   So It seems likely that my idea of a cool  ENSO neutral to weak La Nina seem likely. If we do see a weak La Nina I expect it would reverse back to ENSO neutral for late winter. So a cool neutral or weak La Nina for November into the end of December or so, then we should see the La Nina reverse as we get closer to late winter.

There has been a big change in the ENSO from this Spring compared to now.

Here is a look at the mid-April International Research Institute for Climate and Society (IRI) chart.

Here is a look at the latest International Research Institute for Climate and Society (IRI) chart.

If we do end up with a weak La Nina; I do think the La Nina will be leaning toward the east based idea.

I have to remind you that many other factors are involved besides just the ENSO and La Nina.

The Indian Ocean Dipole is neutral

The Quassi-Biennial Oscillation (QBO):

The QBO started trending easterly back in April. So it looks like a moderate to strong east based QBO is likely. During last year's winter of 2016-2017 we had a La Nina with an west based QBO. So an negative east based QBO would have a completely different impact on our winter. When we have an eastern based QBO we tend to see a warmer stratosphere over the northern polar region, as opposed to a cooler stratosphere when the QBO is west based. Looking at past east based QBO years, the East does tend to see more troughiness.  

The Pacific Decadial Oscillation (PDO):

As it has been for the last few winters, The PDO remains positive. But because of the double La Nina, it is weaker than it was last year. The positive phase of the PDO involves warmer than average water temperatures off the West Coast of North America. The cool negative phase is the opposite.

The PDO has a big impact on where ridging and troughing sets up.  The more positive the PDO the more persistent and deeper the western ridge. I think the PDO will stay a weak positive into the Spring of 2018, where as it would go neutral. Being weaker than last year and the year before that. Will mean the western trough and eastern trough won't be as strong and dominate has they have been. This would have a impact on the storm track and how far south the cold extends.

There are many other teleconnections that I looked at. But I don't want to overwhelm Y'all with all of that data. More or less they support my idea on how winter will turnout.

It is important to remember, how a season turns out is based on a mixture of different teleconnections and other patterns, not the individual teleconnections themselves.

Solar Activity:

This Winter will feature low solar activity. There is a correlation between winter temperatures and sun spot activity.  Years with  low solar activity tend to run colder with more of a negative AO and negative NAO, than winters with high solar activity. Winter's that experience low solar activity also tend to see higher increases of higher latitude blocking around Greenland and Iceland. Which is good if you like colder weather.

Here are a couple of charts that show how all of this characteristically works out .

Here are a few images that show  Ap solar activity and NAO index values.

Here is one from, International Climate and Environmental Change and Assessment Project (ICECAP).

You can see there is a coloration between solar activity and the NAO and AO

When we have negative NAO and AO index values we tend to see more outbreaks of cold air moving out of Canada do to blocking over Greenland.

The Hurricane Season:

This year we have seen a lot of low pressure over North America. This happened when prevailing high pressure set up in the Atlantic. We have seen a lot of high pressure over eastern Canada into the Northeast. So with lower pressure over the Southeast into Caribbean and Gulf land falling tropical cyclones were going to be a given.

I have found that there is a weak association between Atlantic and Pacific tropical activity and winter snowfall.  It all comes down to Accumulated Cyclone Energy (ACE). ACE measures the strength and duration of individual tropical cyclones.

As of Hurricane Olivia this seasons ACE sits at 212. An average year has an ACE of 111. OK but how does this impact our winter? The atmosphere is always seeking a balance. When I looked back at years that had above average ACE. The years that came up were 1950, 1969, 1980, 1989, 1995, 1996, 1998, 1999, ,2000, 2001, 2003, 2004, 2005, 2008, 2010, 2011,  2012. Over 60% of the time, When ACE was above average, the Northeast saw above average snowfall overall, while the Southeast saw below average snowfall overall.

 The ACE data set is small. so finding meaning from extrapolated data is hard.  I don't claim to have found the only answer, because other issues could be involved. But I believe further research should be done in this area. There is also a correlation between ACE and the AMO, which I won't go into too.  

Going into Winter: 

October, November, and into the first part of December will see temperature swings back and forth. That doesn't mean we won't see cold, it just means it won't be really long lasting or widespread. We could see a decent cold outbreak in a couple of weeks. But I do expect to see some early season snows end of October or November, especially in the Snow Belts and higher elevations of New York State and Northern New England. Based on the analogs, there is a good chance for an end of October into November big storm. 

The real winter cold will most likely hold off until around Christmas into  January. Making January, February, into March feeling the most winter-like. Mid to late winter should see more longer lasting cold in the Great Lakes and Northeast.

So, October is going to be very transitory in nature. The overall temperatures in October will be above average. The Northeast has been very dry the last three to four weeks. But Nate is going to bring some much needed rain. So precipitation wise October will end up close to average.

November will see the pattern flip to more in the way of troughing, so we should end up with overall temperatures being  below average, how much is uncertain. But we will have quite a bit of cold locked in Canada.  Unless we see another tropical system, It will also be rather dry, (unless we see something tropical).

December should see temperatures also below average with the end of December ending up most likely well below average.

Northern Hemisphere snowpack:

Snow has been falling in Siberia and we've had a few early season snowfalls in the Rockies into Montana. The models are bullish on the idea of snow in Siberia into China over the next two months.

Here is a look at the snow and ice extent.

With snows flying in Siberia, the more snowpack there is in Siberia and northern Canada, the better the chance for high latitude blocking.  So there is a good chance for negative NAO and a negative AO.   

Wild Cards:

Pacific Jet. for winter 2016-2017 we had a zonal (west to east) flow due to a strong Pacific jet. This is the main reason for the warmth  and wild swings we saw. There is a chance we could see a strong pacific jet again this winter.

There would be cold air in Canada and into the middle of the CONUS (Continental U.S.). depending on how the polar jet acts it could overwhelm the pattern during the last half of winter.  

Blocking, How much we see and where it sets up is as of yet, a little unclear. But I don't think the warmth of this fall will last all the way through winter 2017-2018.


Where the prevalent ridge sets up will be key, as to where the core of the cold sets up. I do expect to see some level of a southeastern ridge. But I don't think it will be as prevalent as last winter. why? Because I think we will see more Greenland Blocking than we saw last Winter.

While the core of the winter cold looks to setup in the Great lakes into the Plains. The Northeast and Mid Atlantic can expect a chilly winter especially from mid winter to end of winter. 

The Southeast is going to see a very warm Winter, that combined with cold air invading the Rockies into the Great Lakes, will allow for plenty of opportunities for wintertime precipitation.  Blocking is going to be a wildcard for winter 2017-2018. But it looks to setup so that the Midwest and Great Lakes will be quite cold, with the Northeast not getting quite that cold. The analogs I looked at, show the storm track should be fairly active. So I don't think snowfall will be all that bad here in the Northeast into the northern Mid Atlantic. An active clipper and Colorado Low storm track will bring many chances for lake effect snows downwind of  Lakes Erie and Ontario. But, the water temperatures in the Great Lakes are warmer than last winter, so unless we can get some really cold air toward the end of November, lake effect snow  could be less than last year.


Temperature profiles in the Northeast are going to be very tricky this winter. This year we will have an east based La Nina. Typically La Nina winters tend to be colder in the Northwest, Central US, and the Great Lakes. The Southeast tends to be quite warm. The Northeast and the Mid Atlantic tend to be in-between somewhere. The strongest widespread cold will be mid it late winter. November will feature some cold . But, I don't expect to see  longer lasting cold for the first third of winter. But end of December thru March have the potential to be cold.

Here are my temperature ideas.


During an La Nina, the typical precipitation pattern sets up as above average over the Northeast. So, snowfall and mixed events for much of New York State and northern New England should be above average. With the Pennsylvania, southern New England, and  Mid Atlantic seeing  average to slightly above average, there is a bigger risk for mixed events in the Mid Atlantic region. The relatively warm waters off the East Coast could allow for a few bigger snowstorm events.

Here are my Northeast snowfall Ideas.

If I have any changes to make, I will post on that around mid November.

Wednesday, August 30, 2017

Was Harvey a result of Climate Change?

On my Facebook weather page, I've been fielding a few questions  from some of you....on Harvey and Climate Change. Y'all have seen the media and others taking about how Harvey was caused by climate change.  The talk is that human caused global warming has lead to weaker prevailing winds, along with the jet stream tracking farther north. Whoever the media is talking to is saying, that these are the changes that caused Harvey to drift around and stall over southeast Texas, leading to the catastrophic flooding we've see in Texas. These people are also saying, In all of U.S. history there has never been a hurricane like Harvey.

My answer to this is baloney.            

I've talked about the gobbledygook that two sides in the climate debate are using, and how they use it to twist the science and fact one way or the other.  Both sides are using junk science.   Many times, I've stated my problems with a lot of the manmade climate change agenda.    I'm not saying that human activities don't contribute to climate change. But I believe cyclical patterns and natural variations have played a much bigger role than human activities. But I won't go down the path of stating my objections to human caused global warming, in the current climate debate, I've said that all before.

As for climate change causing Harvey to act the way he did. There is nothing to support that idea.

We've seen hurricanes bigger than Harvey, we've seen typical cyclones loop and meander around  many times in past years, we've seen tropical cyclones drop enormous amounts of rain before, As in everything to do with weather, it has all happened before. If none of these other times weren't caused by human caused climate change. Then why is Harvey a result of climate change?

.  A couple of years back we had Hurricane Patricia, her winds were 200- 225 mph before weakening before landfall.  It has been Sixty-one years since the first and only Hurricane Hunter plane crash in the Atlantic Basin, brought down by a tropical system it was investigating. That tropical system was Hurricane Janet in Sept of 1955. We had the Great Galveston Hurricane, the Florida Keys Labor Day Hurricane of 1935, The Great New England Hurricane/Long Island Express in 1938, the Great Miami Hurricane in 1926, I can go on and on.



Both of the above came from Weatherbell.
The Galveston Hurricane in1900 was the deadliest hurricane to ever hit the United States and caused between 8000 and 12000 deaths. The same people who say Sandy was a freak storm caused by climate change, seem to forget all about the September 15, 1944 Great Atlantic Hurricane. Just as they forget: Hurricane Camille in August of 1969, Hurricane Hugo in September 1989, Hurricane Audrey in 1957,  Hurricane Andrew in1992, Hurricane Flora in 1963. Hurricane Flora dropped 100.4 inches of rain when it meandered over Cuba. Hurricane Ginny of October 1963


 I can post many other images of hurricanes meandering and doing strange things.
Still not far enough in the past....well we had: Colonial Hurricane of 1635, The Great Chesapeake Bay Hurricane of 1769, The Great Coastal Hurricane of 1785, there are dozens of others.

I've talked about the great East Coast Hurricanes of the 1950's and 60's, with names of Hazel, Connie, Diane, and Donna. So I ask again if none of these were caused by global warming and climate change then why is Harvey?