Summer forecast for North America -- severe heat waves ahead

[Roger Smith]

This can be a very short long-range forecast. I have run my developmental research model using all the data available and the results look very much like the past two years for the northeast U.S., Great Lakes and Ohio valley, but this year, I expect the heat to be very widespread as a very anomalous bulge in the upper circulation takes up semi-permanent residence over the eastern half of the U.S. with a secondary ridge on the west coast. Temperature anomalies are likely to be close to record values (other than perhaps 1936) in central regions as well as extending to the east coast.

June could be somewhat closer to normal than July and August at least in the northeast, but it will be a case of widespread severe heat building up from the central plains into the east and then the pattern locking in until mid to late September.

Rainfall is likely to be considerably below normal in many parts, but I think it will be an active severe storm season in the northern plains, upper and central Great Lakes into upstate New York and northern New England, southern Quebec and New Brunswick.

There is likely to be heavy rainfalls in parts of Montana and Alberta, extending into parts of Wyoming and Idaho, and eastern B.C., while the west coast should have a detached regime of warm, dry weather from a ridge near the Gulf of Alaska. Only a weak upper low will influence the Great Basin and the desert southwest will also be above normal in temperature with possibly a very active monsoon season.

My outlook calls for a very active North Atlantic tropical season with a predicted total of 21 named storms, 13 hurricanes and 7 major hurricanes. All theatres will become active at times. With this level of storm development, there would almost certainly be 2-5 significant landfalling storms for the U.S. and/or eastern Canada.

I might add some maps but would think that this pattern pretty much explains itself. Not surprisingly, the output indicates a peak in the severe heat conditions in late July and early August, a time of year when record highs can easily surpass 105 F and reach 110 F.

A forecast has been prepared for Europe and this shows widespread heat there also, trending closer to normal in the U.K. and Ireland but still somewhat on the warm side, once the current chilly blocking pattern totally breaks down in about a week or two. Remnants of this blocking may retrogress into eastern North America leading to a few relatively cool days for the east before this heat wave situation intensifies.

[andyhb]

My outlook calls for a very active North Atlantic tropical season with a predicted total of 21 named storms, 13 hurricanes and 7 major hurricanes. All theatres will become active at times. With this level of storm development, there would almost certainly be 2-5 significant landfalling storms for the U.S. and/or eastern Canada.

That seems a little sketchy given that virtually all of the other forecasts out there call for a quiet season. That is approaching 2005 numbers with that call.

[carley5]

Any ideas on hurricane landfall locations?

[Mr Torchey]

Good luck, your winter forecast was certainly one of the best of the year!

Bring on the heat!

[Ginx snewx]

AAM

[CoastalWx]

I'd like to know how you came up with 21 named storms.

[nzucker]

And here's the beginning:

[easternsnowman]

Good luck, your winter forecast was certainly one of the best of the year!

Bring on the heat!

He did miss feb big time.

[Mr Torchey]

He did miss feb big time.

yep but still, better than most by a long shot

[phlwx]

That seems a little sketchy given that virtually all of the other forecasts out there call for a quiet season. That is approaching 2005 numbers with that call.

with a developing Nino no less.

[Tropical]

A+ forecast. I like the reasoning.

[uncle W]

here is the summer map for developing el nino years...2002 was probably the hottest of the bunch...1991 was hot also...2004 and 2009 were quite cool for the northeast...

[Roger Smith]

I went high on the NATL tropical count in part because 1933, 1936 1995 and 2005 were in the analogue set. I factored in modern observational ability and figured this could head into near-record territory. Landfall ideas would be pretty vague but the Gulf coast has been conspicuously quiet for several years, and you have to think eastern Florida would be near a rainfall positive anomaly in the kind of pattern being discussed.

[donsutherland1]

He did miss feb big time.

Overall, he did a terrific job in terms of the forecast as a whole. In addition, his seasonal snowfall forecast was outstanding.

[Harry]

This can be a very short long-range forecast. I have run my developmental research model using all the data available and the results look very much like the past two years for the northeast U.S., Great Lakes and Ohio valley, but this year, I expect the heat to be very widespread as a very anomalous bulge in the upper circulation takes up semi-permanent residence over the eastern half of the U.S. with a secondary ridge on the west coast. Temperature anomalies are likely to be close to record values (other than perhaps 1936) in central regions as well as extending to the east coast.

June could be somewhat closer to normal than July and August at least in the northeast, but it will be a case of widespread severe heat building up from the central plains into the east and then the pattern locking in until mid to late September.

Rainfall is likely to be considerably below normal in many parts, but I think it will be an active severe storm season in the northern plains, upper and central Great Lakes into upstate New York and northern New England, southern Quebec and New Brunswick.

There is likely to be heavy rainfalls in parts of Montana and Alberta, extending into parts of Wyoming and Idaho, and eastern B.C., while the west coast should have a detached regime of warm, dry weather from a ridge near the Gulf of Alaska. Only a weak upper low will influence the Great Basin and the desert southwest will also be above normal in temperature with possibly a very active monsoon season.

My outlook calls for a very active North Atlantic tropical season with a predicted total of 21 named storms, 13 hurricanes and 7 major hurricanes. All theatres will become active at times. With this level of storm development, there would almost certainly be 2-5 significant landfalling storms for the U.S. and/or eastern Canada.

I might add some maps but would think that this pattern pretty much explains itself. Not surprisingly, the output indicates a peak in the severe heat conditions in late July and early August, a time of year when record highs can easily surpass 105 F and reach 110 F.

A forecast has been prepared for Europe and this shows widespread heat there also, trending closer to normal in the U.K. and Ireland but still somewhat on the warm side, once the current chilly blocking pattern totally breaks down in about a week or two. Remnants of this blocking may retrogress into eastern North America leading to a few relatively cool days for the east before this heat wave situation intensifies.

A+ forecast. I like the reasoning.

ummmm.... What reasoning?

here is the summer map for developing el nino years...2002 was probably the hottest of the bunch...1991 was hot also...2004 and 2009 were quite cool for the northeast...

Note that in MOST CASES ( exception i believe was 1958? and 2001 )going back to 1950 when we had a Nina the winter before and transitioned to either neutral or Nino we ended up near normal or below normal in the GL/NE. In cases like 1991 and 2002 we went from neutral ( NOT Nina ) to El'Nino. He is really betting against the odd's i would say. JMHO Again i am talking summer as a whole and thus June, July, and August.

See 1951, 1965, 1972, 1976, 1986, 1989, 1996, 2009

[Roger Smith]

Thanks Don ... and in response to several others, I have posted details on my research methodology in the past, and this remains the basis for the current forecast. It is a theoretical model based on assumptions about interactions between our atmosphere and the solar system magnetic field structure. Teleconnections are assumed to fall into the more statistically-driven numerical output, but as I believe the winter forecast demonstrated, the technique does not carry assumptions about higher probability outcomes of Pacific or other signals. This could be a weakness or a strength over a long period of forecasting, but I keep searching for additional components to add to the model.

To give some idea of how stacked my deck is for warmth, all of these years are in the analogue set -- 1916, 1918, 1921, 1933, 1936, 1948, 1955, 1973, 1975, 1987, 1988, 1995, 2005, 2006. That is not quite the full set and there were a few cooler summers such as 1965 in the data. The forecast is generated by weighted blends of individual sector analogues, adjusting dates slightly to fit the field structure -- the tolerance is something like 5% of the full range of annual data for each component. Example, if I am looking for "J-field" analogues, I take those which would provide similar components to July 15 of this year from all cases in July (1-31) in past data and move the data to line up with the postulated effect. Then I check against a wider cross-section to investigate changes over time. I also weight more recent data somewhat higher.

I was a bit concerned about 1965 being in the data set, so I had a close look at its full range of model parameters, it gets in from two factors out of about twelve, and most of the other ten were colder signals than this year should provide. So I was satisfied that it wasn't showing a weakness in the factors that placed it in the data set. If the summer looks more like Uncle W's main map (and not the 91-02 map) then I may have to revisit that reasoning.

Well anyway, the numerical output runs about as high as it did through Dec to Jan for the winter forecast, and I think the February glitch was partly feedback, as there was a weak attempt to build a trough in late February and even some wintry weather in New England. Although this technique is basically objective-statistical (driven by data and not reasoning) I do have some experience now with the process of tracking field components and studying how they interact when they appear in a weather pattern together. In this case, the warming influences should reinforce well, then also, I factored in a data component derived from similar solar background conditions and that also shows an upward trend.

Seems that the NHC have given my tropical forecast a "mulligan" with Alberto ... thanks for that, I guess that makes it 20 more to go.

[Harry]

Thanks Don ... and in response to several others, I have posted details on my research methodology in the past, and this remains the basis for the current forecast. It is a theoretical model based on assumptions about interactions between our atmosphere and the solar system magnetic field structure. Teleconnections are assumed to fall into the more statistically-driven numerical output, but as I believe the winter forecast demonstrated, the technique does not carry assumptions about higher probability outcomes of Pacific or other signals. This could be a weakness or a strength over a long period of forecasting, but I keep searching for additional components to add to the model.

To give some idea of how stacked my deck is for warmth, all of these years are in the analogue set -- 1916, 1918, 1921, 1933, 1936, 1948, 1955, 1973, 1975, 1987, 1988, 1995, 2005, 2006. That is not quite the full set and there were a few cooler summers such as 1965 in the data. The forecast is generated by weighted blends of individual sector analogues, adjusting dates slightly to fit the field structure -- the tolerance is something like 5% of the full range of annual data for each component. Example, if I am looking for "J-field" analogues, I take those which would provide similar components to July 15 of this year from all cases in July (1-31) in past data and move the data to line up with the postulated effect. Then I check against a wider cross-section to investigate changes over time. I also weight more recent data somewhat higher.

I was a bit concerned about 1965 being in the data set, so I had a close look at its full range of model parameters, it gets in from two factors out of about twelve, and most of the other ten were colder signals than this year should provide. So I was satisfied that it wasn't showing a weakness in the factors that placed it in the data set. If the summer looks more like Uncle W's main map (and not the 91-02 map) then I may have to revisit that reasoning.

Well anyway, the numerical output runs about as high as it did through Dec to Jan for the winter forecast, and I think the February glitch was partly feedback, as there was a weak attempt to build a trough in late February and even some wintry weather in New England. Although this technique is basically objective-statistical (driven by data and not reasoning) I do have some experience now with the process of tracking field components and studying how they interact when they appear in a weather pattern together. In this case, the warming influences should reinforce well, then also, I factored in a data component derived from similar solar background conditions and that also shows an upward trend.

Seems that the NHC have given my tropical forecast a "mulligan" with Alberto ... thanks for that, I guess that makes it 20 more to go.

What makes them such great analogs though?

[GaWx]

I went high on the NATL tropical count in part because 1933, 1936 1995 and 2005 were in the analogue set. I factored in modern observational ability and figured this could head into near-record territory. Landfall ideas would be pretty vague but the Gulf coast has been conspicuously quiet for several years, and you have to think eastern Florida would be near a rainfall positive anomaly in the kind of pattern being discussed.

1) I'm wondering if 1933 and 1995 should be in the tropical analogs since they were La Nina's during the heart of the season and there's no indication of a La Nina returning immediately.

2) Keeping in mind the possibility of a developing El Nino (still in doubt of course), your analogs that are El Nino's averaged only 8 NS.

[cyclogent]

I can see this working for the month of June being hot for the Apps westward but after that the warming of ENSO SSTs argues for average or slightly below average temps

[jdrenken]

He did miss feb big time.

Compared to the constant blue-raspberry kool-aide being served...he was on the money.

[Ottawa Blizzard]

Quite a different forecast from JB who, jusging from his tweets, seems to be hinting at a cool summer along the lines of 1976. He's been busting so far, though given May is turning out quite warm, at least in Ontario.

[Brian5671]

Quite a different forecast from JB who, jusging from his tweets, seems to be hinting at a cool summer along the lines of 1976. He's been busting so far, though given May is turning out quite warm, at least in Ontario.

JB has gotten zilch right in the past 9 months outside of the October snowstorm....

[PSUBlizzicane2007]

My outlook calls for a very active North Atlantic tropical season with a predicted total of 21 named storms, 13 hurricanes and 7 major hurricanes. All theatres will become active at times. With this level of storm development, there would almost certainly be 2-5 significant landfalling storms for the U.S. and/or eastern Canada.

There is absolutely no evidence for such an outlook to be made.

[Ottawa Blizzard]

JB has gotten zilch right in the past 9 months outside of the October snowstorm....

He's grasping at straws trying to find or manufacture proof that we're in for a significant global cooling spell similar to the late 70s.

[easternsnowman]

Overall, he did a terrific job in terms of the forecast as a whole. In addition, his seasonal snowfall forecast was outstanding.

I believe he called for a cold snowy Feb which seemed to indicate that snowfall will be at least near average for the season. In the cities it was generally around 25% or so of average.

[jdrenken]

I believe he called for a cold snowy Feb which seemed to indicate that snowfall will be at least near average for the season. In the cities it was generally around 25% or so of average.

You honestly can't make that assumption that "cold and snowy" would make up for the departures from earlier in the year.

Again...despite February's call being a bust, his forecast as a whole did quite well in the Lower 48.

[Guru Of Reason]

No offense, but I hope you bust HARD.

[weatherwiz]

I believe he called for a cold snowy Feb which seemed to indicate that snowfall will be at least near average for the season. In the cities it was generally around 25% or so of average.

I don't think it's completely fair to judge someone's forecast based on their precipitation forecast, especially when it comes to snowfall. Forecasting precipitation/snowfall is much more of a crap shoot than forecasting temperatures in the long-range. When it comes to snowfall one or two storms can make or break a forecast.

[ptb127]

Hope his temp forecast for the US is correct. I don't follow the tropics much but it seems that's where many of you have a problem with his fcast

[Roger Smith]

Can't add much to previous comments, just wait and see time now. Tropical forecast is correlated with temperature forecast, so I suspect it all goes one way or the other.