This might take a couple posts to actually get through the whole picture. It seems like there are 2 dominate issues and maybe a couple minor issues that are going to guide our fall and winter.
I’m a big believer that hurricanes can effect our weather via two mechanisms. First, they can deliver moisture to across Mexico from the Gulf of Mexico and the Caribbean Sea to our doorstep. This can cause precipitation to amounts to be higher than average. Second, they can actually disrupt the tropical moisture flow and locations of dominant high pressure systems. They can actually draw the moisture from thousands of miles around, drying out Northern Arizona. One of the big contributors to hurricane formation is the water temperatures in the Caribbean and Gulf of Mexico. La Nina’s are typically more favorable for active hurricane seasons. El Nino’s tend to generate a strong flow in the opposite direction of the normal hurricane paths, preventing hurricane development.
This Atlantic Hurricane Season is forecasted to be fairly strong with 14-20 named storms and 8-12 hurricanes (Climate Predictionn Center Hurricane Outlook 8-5-2010). This outlook was just updated. Although the season has started slowly, August to October are the peak months. There is still time for a strong season.
So, La Nina and El Nino can alter weather patterns significantly. El Nino in the fall and winter usually means wetter conditions. Last winter, an El Nino episode provided us with nearly 6 feet of snow in one week. Currently, the Pacific Ocean is building a La Nina episode. It just recently started to develop and forecasts vary as to how deep it will be. When looking at La Nina’s as an input to the precipitation outcome for the August-October time frame, There have been 7 La Nina’s through these months in the last 25 years, 1985, 1988, 1995, 1998, 1999, 2000, 2007. It’s arguable that 2000 was not a La Nina, but I have included it because the region was cold and clearly met the La Nina definition (El Nino Sea Surface temperature anomaly <-0.5) on either side of August-October.
| Year |
La Nina SST Anomaly |
| 1985 |
-0.5 |
| 1988 |
-1.3 |
| 1995 |
-0.5 |
| 1998 |
-1 |
| 1999 |
-1 |
| 2000 |
-0.4 |
| 2007 |
-1 |
In 1995 and 2007, the La Nina’s were just starting around the mid to late summer time period. For this year, the outlook for the August-October time frame is between -0.5 and about -1.25. Again, this could match either 1995 or 2007. Notice a theme here? Don’t get too bent on 1995 and 2007 being good predictors quite yet.
Summary of ENSO Model Forecasts
Another interesting sea surface temperature anomaly to track is the Pacific Decadal Oscillation. It’s a broader, longer term look at the Pacific and how it behaves. It has been negative for most of the last 3 years. With El Nino ending, it returned to negative in June. This is expected to last, and intensify for quite sometime. We can assume it will be negative for the next 6 months. How do the other years look?
| Year |
PDO |
| 1985 |
0.51 |
| 1988 |
-0.09 |
| 1995 |
0.61 |
| 1998 |
-0.94 |
| 1999 |
-1.57 |
| 2000 |
-1.24 |
| 2007 |
-0.44 |
1985 amd 1995 were both positive. But, 2007 and 1998 were negative.
Now, the hurricane picture in those La Nina years is interesting, too.
| Year |
Atlantic Hurricanes |
Atlantic Tropical Storms |
Gulf Hurricanes |
Gulf Tropical Storms |
| 1985 |
7 |
3 |
4 |
5 |
| 1988 |
5 |
12 |
3 |
5 |
| 1995 |
11 |
19 |
4 |
6 |
| 1998 |
10 |
14 |
2 |
5 |
| 1999 |
8 |
12 |
1 |
2 |
| 2000 |
8 |
15 |
2 |
4 |
| 2007 |
6 |
15 |
2 |
4 |
Remember, 14-20 named storms and 8-12 hurricanes are in the outlook. In this chart, tropical storms are equivalent to named storms. 1995 continues to be a potential match for this year. 2007 falls short on the total number of hurricanes, but is in the zone for total named storms. 1998 and 2000 are in the zone as well. The numbers of Gulf Storms are higher in many of the earlier years. But, 2007 still had 4 named storms in the Gulf of Mexico.
Here is where the real conundrum kicks in. When you take all of this data and add Flagstaff’s August to October precipitation to the chart you get this:
| Year |
Flagstaff Precipitation |
La Nina SST Anomaly |
PDO |
Atlantic Hurricanes |
Atlantic Tropical Storms |
Gulf Hurricanes |
Gulf Tropical Storms |
| 1985 |
7.15 |
-0.5 |
0.51 |
7 |
3 |
4 |
1 |
| 1988 |
6.15 |
-1.3 |
-0.09 |
5 |
12 |
3 |
2 |
| 1995 |
3.88 |
-0.5 |
0.61 |
11 |
19 |
4 |
2 |
| 1998 |
10.38 |
-1 |
-0.94 |
10 |
14 |
2 |
3 |
| 1999 |
6.99 |
-1 |
-1.57 |
8 |
12 |
1 |
1 |
| 2000 |
7.04 |
-0.4 |
-1.24 |
8 |
15 |
2 |
2 |
| 2007 |
5.84 |
-1 |
-0.44 |
6 |
15 |
2 |
2 |
The average precipitation for the August to October time periods in all these La Nina years is 6.78 inches, with a standard deviation of 1.95 inches. This put 1995 and 1998 at opposite ends of the spectrum and outside one standard deviation of the average. The positive PDO in 1995 with a La Nina, and an active hurricane season seems to point to a dry period.
To me, this seems to me to indicate that 2007, and maybe 1998, could be the best analogs for this year. A couple caveats about 1998. La Nina started just a couple months earlier than this year. It also followed a strong El Nino, in fact the strongest.
So, I am going to pick 1998 as the analog year for my forecast. Which would mean that we should see greater than average precipitation(1984-2009 average is 7.61 inches) for these months. In 1998, Flagstaff received 3.32 inches in August, 4.76 in September and 2.96 in October. I think we will repeat a wetter than average trend similar to this, not 1995′s very dry trend or 2007 average trend.
But what about winter? La Nina means drier…or does it?
