Energy Management

Q&A: September Winter 2017-2018 Weather Outlook Webinar

4 min read

Last month we hosted a special edition Winter Weather Forecast Webinar. Constellation’s meteorologist, Andrew Durante, gave a 30-minute in-depth overview of the 2017/2018 winter forecast. Attendees learned about the current winter 2017/2018 forecast and how it could affect energy markets and energy strategies; key drivers behind market forecast and how they link to the current winter forecast; winter forecasts from years past as well as how cold previous winters were after active hurricane seasons.

We received a few questions from attendees. Please check out the answers below to see insight into the questions attendees had: 

1. Do the recent Atlantic Coast hurricanes and warm water have an impact on winter weather predictions?

None that has been proven. Some studies have shown no statistical significance between Atlantic Coast hurricanes and winter temperatures. If anything, the southern United States showed a slight warmer signal; however, that may have more to do with La Nina.

2. In the summer, colder temperatures mean less power demand. In the winter, colder temperatures mean more power demand. When do you think the transition point on how colder temperatures impact prices occurs?

The transition point tends to occur sometime in early November. October and November are referred to as “shoulder” months due to temperatures averaging closer to 65 degrees F, which is a threshold for heating/cooling demand. It is the buildup of overnight heating load (overnight temperatures get down consistently into low 40s) in November that begins the heating season.  We typically see this play out in the weekly U.S. Energy Information Administration (EIA) storage report numbers each Thursday when withdrawals start to exceed ~20 Bcf/week. If you’re interested in receiving these weekly reports from Constellation, subscribe here.

3. Was the Quasi-Biennial Oscillation (QBO) not high for both the polar vortex and last winter – two very different winters?

Yes, the QBO was high for the polar vortex winter and last winter. However, there were major differences across the northeast Pacific. During the polar vortex winter, extremely warm water temperatures were located across the Gulf of Alaska (very positive Pacific Decadal Oscillation (PDO)). This maintained very strong blocking over Alaska and allowed very cold air to spill into the Midwest and East. The polar vortex winter actually had very little blocking over the Atlantic. Cold water temperatures over the Gulf of Alaska prevented any Alaskan blocking for last winter.

4. How much has demand dropped in comparison to supply in the same period shortly after Hurricane Harvey?

Hurricane Harvey’s impact was greater on demand than supply. On the production side, per EIA’s weekly Thursday report, production declined from 72.8 Bcf/day for the week ending August 23rd to 71.7 Bcf/day for the week ending August 30th. Dry production recovered to 73.4 Bcf/day for the week ending September 6th, climbing above pre storm levels.

On the demand side, there were widespread power outages in ERCOT, which reported a loss of load of approximately 15-20 GW (normal load is 65,000 MW and it declined to about 45,000 MW). The biggest impact was to gas-fired generation demand. It declined nationally from 35.6 Bcf/day for the week ending August 23rd, down to 29.4 Bcf/day for the week ending August 30th and remained flat for the week ending September 6th. Exports to Mexico declined from 4.4 Bcf/day to 3.9 Bcf/day as work crews at compressor stations in South Texas, along the Tennessee Gas pipeline, were evacuated. Pipeline receipts into Sabine Pass fell from 1.8 Bcf/day down to 0.5 Bcf/day for week ending September 6th as liquefied natural gas (LNG) storage tanks reached capacity.

5. Historically, how accurate has the El Nino Southern Oscillation (ENSO) and QBO indexes been?

Predicting an El Nino and La Nina is relatively skillful; however the strength of any event still carries low skill. The QBO is relatively easy to predict month-to-month, however week-to-week effects that the QBO has on blocking tend to be less skillful.

6. Any preliminary look on precipitation/snowfall for the winter?

Precipitation could be above normal across the northern tier (Upper Midwest and Northeast) and below normal for the Ohio Valley and Mid-Atlantic. Below normal precipitation is expected for Texas and the Southeast. The Northwest could see an above-normal hydro year, however California may see a drier-than-normal year.

September Winter Webinar 1.pngSource: NOAA

7. With current Artic ice levels on the low-side (below average), does low Artic sea ice correlate to a warm winter?

The correlation is relatively weak. We have only had reliable satellite data over the Arctic since 1979, so relatively speaking there are not many years to have confidence on a strong correlation between Arctic ice and U.S. winter temperatures.

Below, the graph shows years where there were below-normal Arctic ice conditions in September. The years included 1999, 2000, 2002-2017.

Below-Normal Sea Ice Years: 
September Winter Webinar 2.pngSource: NOAA

The resulting temperature map for all of the following winters show slightly above-normal temperatures throughout most of the nation, with a few pockets of below-normal temperatures for the Midwest.

Resulting Temperatures:
Sept QA Blog Image 3.png


The below ice graph shows years where there was above-normal Arctic ice in September. The years included 1979-1989, 1992, 1994, 1996, 1997 and 2001.

Above-Normal Sea Ice Years:
September Winter Webinar 4.png

Source: NOAA

The resulting temperature map below for all of those years show a slightly colder map, especially across the West and South. The Midwest is actually a bit warmer versus the low ice years; however most of the nation shows colder signals with higher sea ice.

Resulting Temperatures:

September Winter Webinar 5.png

Source: Earthstat

If you missed our September webinar, you can listen to a recording here. Our next Energy Market Outlook webinar will be held on Wednesday, October 18th at 2 PM EST. You can register for our October webinar here.



You may also be interested in these related articles: