Very Early Look at Flood Prospects

Winter 2009/2010 Outlook …and Possible Spring 2010 Flood Impacts for Western and Central North Dakota
 
Given the severity of spring 2009 flooding across North Dakota, there are many people who are already contemplating what spring 2010 will bring to western and central North Dakota with respect to flooding. This is nearly impossible to answer with any degree of certainty because no data exist on many of the factors involved in spring flooding. These factors in no specific order of importance are: 1.) available surface storage in dams, ponds or wetlands, 2.) soil moisture and temperatures at time of melt, 3.) liquid water equivalent in the snowpack at time of melt, 4.) rate of melt, 5.) extra precipitation received during the melt, and 6.) the potential for ice jams that significantly affect stream flow. However, as we exit an unusually warm and dry November throughout the western and central parts of the state, and after a cold and wet October, the question can at least be addressed qualitatively in regard to the conditions of the past spring’s flooding.
 
Available Surface Storage of Runoff: In regard to available surface storage of spring runoff, it is fairly safe to say that most small reservoirs, livestock dams, wetlands, and ponds are full or nearly full throughout the state. The spring flooding filled them to overflowing in many instances, the cool and wet summer kept them fairly full, and the October rains and snow topped them off.  This is in sharp contrast to the past spring where much of the state was coming out of a prolonged drought, which had lowered nearly all surface water bodies in the western and central part of the state. In fact, some large wetlands were dry prior to this spring and others were mere mudholes. 
 
Soil Moisture Conditions: With the unusually warm and dry November, it is still a little too early to tell what the soil moisture conditions will be next spring. We often have solidly frozen ground by now, which prevents infiltration of water that can freeze in the surface soil and produce a nearly impermeable ground surface. When combined with an early melt this tends to generate a very high ratio of runoff from the snowmelt compared to infiltration. However, the dry month of November has substantially lowered the excess moisture present in surface soils and the near to mid-term forecast does not call for much moisture. If this forecast verifies, then soils will freeze by mid-December with normal to below normal moisture levels across the western part of the state. This, again, is in sharp contrast to the fall of 2008 where an early November storm thoroughly saturated much of the state just before soil temps fell below freezing.  A map of current soil moisture with soil moisture values displayed as a percentile ranking is given below. A ranking of 70 percentile on the map indicates that 30 percent of the years on record had more soil moisture and 69 percent of the historic record are comparatively drier. As seen in the map, more extreme southern and eastern parts of North Dakota are still very wet and should not be expected to allow much infiltration if the soils freeze under these conditions.
 
 
Liquid Water Equivalent in the Snowpack: The water equivalent of a snowpack that does not yet exist is impossible to measure or even reasonably estimate at this point. However, we do know that a weak to moderate El Niño pattern has formed and will affect the temperature and precipitation received in the upper Great Plains area, including North Dakota. This strongly suggests that the upcoming winter season will be different than last year’s. During the summer of 2009, an El Niño began developing and has continued throughout the fall. During typical El Niño winters, temperatures for our regionare 4 to 5 degrees Fahrenheit above normal, and snowfall tends to be 67% to 85% of normal. While there are other large scale climate signals that can decrease the affect of the El Niño, the winter of 2009/2010 is expected to be warmer and drier than last year.
 
Classic El Niño winters feature less frequent arctic air intrusions and fewer snow storms. Examples of recent El Niño winters include 2006/2007, 2002/2003, 1997/1998, 1994/1995 and 1991/1992. Below is a table with the measured snowfall in Bismarck, Minot, and Dickinson; the average of the El Niño years and the 30-year average snowfall.
 
 
Station
Season
 
 
 
 
 
6 Season
1971-2000
 
1987/1988
1991/1992
1994/1995
1997/98
2002/2003
2006/2007
average
Normal
Bismarck
31.5
57.5
67.1
26.6
33.2
46.1
43.7
49.9
Minot
14.1
39.5
51.9
52.4
48.2
34.7
40.1
46.2
Dickinson
19.0
21.8
15.9
32.5
13.5
38.5
23.5
35.0
 
Overall, there tends to be almost 6-10 inches less snow during an El Niño winter (December through February). Typically, the snows do not build up as much as the warmer temperatures often melt the snow between storms. This prevents the snow from accumulating to average depths.
 
Accumulating large snowpacks tends to be more difficult during an El Niño because of warmer average temperatures, which often serve to melt snow and can lead to more sublimation. Winter season temperatures tend to be 4-5 degrees Fahrenheit warmer across the Northern Plains during an El Niño. Below is a table of the same 6 seasons, except the December through February average temperatures are compared to normal.
 
Station
Season
 
 
 
 
 
6 Season
1971-2000
 
1987/1988
1991/1992
1994/1995
1997/98
2002/2003
2006/2007
average
Normal
Bismarck
15.9
25.1
15.2
23.9
16.3
15.6
18.7
14.5
Minot
16.1
23.9
14.5
22.9
14.2
16.2
18.0
14.0
Dickinson
15.0
26.1
20.6
23.4
17.5
17.8
20.1
15.7
Temperatures are in degrees Fahrenheit
 
Rate of Melt: Another factor that cannot be determined until just days before, or maybe even during the melt is the rate at which the melt will occur. One could and should reasonably infer that, just like nearly every other year, some areas will experience a much faster rate of melt than other areas. This factor will dictate local flooding and be especially important in deciding which areas suffer overland flooding. For example, southwestern and south central North Dakota had a much faster rate of melt in 2009 than the northwestern and north central part of the state. As a direct result of the melt rate this past spring, some areas with much higher liquid water equivalent in the snowpack did not flood as severely as other areas with a much lower liquid water equivalent.
 
Extra Precipitation Received During the Melt: Not only does extra precipitation add directly to the total amount of runoff, but this is a significant factor in determining the rate of melt as it can substantially speed up the melting process of ice and snow on the countryside. It is a literal impossibility to predict this far in advance, but it can truly create flood conditions on very short notice if all the other necessary pieces of the puzzle are in place.  
 
Ice Jams: Another factor that is mathematically impossible to predict because of its random nature are ice jams. Intuitively though, there are indicators and conditions that are known to increase the likelihood of ice jam related flooding. The conditions that do not change from year to year that can lead to ice jams are those related to the stream channel itself such as meandering streambeds and artificial constrictions, such as bridges, that tend to catch ice and other debris and inhibit the normal stream flow. The conditions that tend to change from year to year which do affect the potential for ice jams are ice thickness and timing of the melt from upper basin to lower basin. High river levels going into the winter freeze-up tend to be able to generate thicker sheets of ice simply because of the deeper water covering larger areas, and more ice equals greater risk. Of course, one must also have the necessary cold temperatures to freeze the ice to a greater thickness. Simply having higher water going into winter does not necessarily equate to more ice, nor does colder winters equal more ice unless the water is deep enough. Some of the ice formed last year on the western rivers in the state came despite the rivers being fairly low going into winter and the amount of ice was fairly surprising even with the cold winter. Much of the ice, especially on the Little Missouri River, can be attributed to a rather anomalous runoff from a rain event in mid-winter that pooled on top of the Little Missouri River and then froze in place on top of the existing ice sheet. This year we are heading into winter with normal to above normal water levels on many streams and this could be a slight indicator of more trouble again in the spring with ice jams than in normal years given relatively “normal” conditions.
 
 
Summary: In general the western two-thirds of North Dakota tends to be in better shape than the eastern third going into this winter simply because of the extremely moist conditions the Red River Valley will likely have going into winter conditions. The available indicators for the rest of the state are a mixed bag with some pointing towards increased risk of flooding (lack of available surface storage) and others are pointing towards a lessened chance of flooding (El Niño, generally dry to normal soil moisture conditions). One thing that is a certainty, though, is that if we were to experience spring melt conditions similar to last year, the lack of available surface storage suggests we will not need the near record snowfall to create similar flooding next spring.


Return to News Archive

USA.gov is the U.S. government's official web portal to all federal, state and local government web resources and services.