Earlier seasonal snowmelt runoff and increasing dewpoints - Upper Midwest
- I am a hydrologist living in Chanhassen, Minnesota. I am one of the
concerned scientists that gave signature approval for:
A Letter From U.S. Scientists
The State of Climate Science: October 2003
I received notice earlier this month that over 1000 scientists gave their
signatures in support of the Letter on climate change.
I am asking scientists that signed the Letter and other people to take a
look at my final draft paper that I will be presenting at the National
Weather Service (NWS) - Climate Prediction Center (CPC) Workshop in Reno,
Nevada: 20-23 Oct 2003.
If you choose to take a look at my final draft, I would appreciate seeing
any comments you may have. I would appreciate your support. At this
date, I am still uncertain on whether or not my employer (NWS North
Central River Forecast Center) is supportive of my effort on this paper.
The narrative portion of my paper is on the Minnesotan's For
Sustainability (MFS) website.
The MFS website is:
The narrative of my paper on the MFS website is called: "Snowmelt &
Dewpoints in Minnesota, Wisconsin, and North Dakota". The URL for my
If you would like to view the figures and tables that go with the paper
please send me an e-mail message to [ npat1@... ].
The figures and tables are in Excel format. I can forward the Excel
spread sheets or I could fax or mail the printouts from the figures (4)
and tables (2).
I believe that Figure 1 on earlier snowmelt runoff for three major
headwater rivers is very telling on climate change.
I would appreciate any comment that people may have on my paper and work.
A copy of the narrative is included below, copied from my paper on the
Affiliation is for identification purposes only
Patrick J. Neuman (Pat)
National Weather Service
North Central River Forecast Center (NCRFC)
Affiliation is for identification purposes only
NCRFC is collocated in Chanhassen, MN (along with the NWS Weather
Forecast Office for east central Minnesota and northwest Wisconsin, and
the National Operational Hydrologic Remote Sensing Center).
The narrative paper that follows is from the website at:
Earlier in the Year Snowmelt Runoff and Increasing Dewpoints for Rivers
in Minnesota, Wisconsin and North Dakota
Patrick J. Neuman, Snow Hydrologist, NWS, NCRFC
September 11, 2003
II. Snowmelt physics
III. Hydrologic area and data sources
IV. Timing of annual snowmelt runoff
V. Average monthly dewpoint
VI. Conclusions on the timing of snowmelt runoff and humidity
VII. Air temperatures
VIII. Additional discussion
Daily river flow data were used to evaluate the timing of snowmelt runoff
at three river stations within the Northern Great Plains and Upper
Midwest. Timing of snowmelt runoff is shown by a X-Y plot using 10 year
moving averages for �Annual Beginning Day of Snowmelt Runoff� at the
river stations for 1910-2003. Average dewpoint plots are shown for three
climate stations near the river stations. The plots show monthly January
- April dewpoint for 10 year moving averages for 1948 to 2003. A
discussion of snowmelt physics is included, describing how humidity as
measured by dewpoint affects the rate of snowmelt. Based on the study
results, shown by the plots on the timing of annual snowmelt runoff and
by plots of dewpoint averages at climate stations, conclusions are
reached, and recommendations are given.
II. Snowmelt Physics
After a long period of cold weather, a snowpack can absorb large amounts
of heat before thaw occurs. Once the temperature of the snowpack reaches
zero degrees Celsius throughout, liquid water starts forming within the
snowpack. When the liquid water exceeds a threshold (about 15 percent of
total snowpack water equivalent), snowmelt begins.
Solar radiation is the dominant energy transfer for snowmelt during clear
sky periods. Usually snowmelt occurs on south facing slopes and hilltops
before snowmelt occurs on north facing slopes and other parts of the
basin. In winter and early spring, sun angles are low and days are
short; thus snowmelt from solar radiation alone during this period is
usually gradual and intermittent.
The significance of latent heat for snowmelt has been described by Dunne
and Leopold (1978):
�If water from moist air condenses on a snowpack, 590 calories of heat
are released by each gram of condensate. This is enough energy to melt
approximately 7.5 gm of ice, which when added to the condensate yields a
total of 8.5 gm of potential runoff�.
Latent and sensible heat transfers can result in high snowmelt rates, as
warm moist air moves into a region. Latent and sensible heat transfers
can cause rapid snowmelt from all parts of a basin simultaneously, day
and night, even during winter. Warm temperatures, high humidity, and
strong winds have large effects on the rate of snowmelt. In comparison,
heat supplied by rainfall is usually minor, unless a warm rainfall of
long duration occurs. A more detailed description of equations for
snowmelt are given by Price and Dunne (1976).
Dunne and Leopold (1978) show that �highest melt rates were associated
with the warm sector of a large weather disturbance� (Quebec, May of
1973). For the last three days of an eight day melt of the snowpack in
May of 1973 (Quebec), melt due to latent heat was shown to be nearly
equal to melt from net radiation, and melt from latent heat during the
last three days was shown to be around 50 percent of the melt due to
sensible heat transfer from atmospheric convection (mixing).
From the theoretical and physical descriptions given above, it is clear
that the rate of snowmelt increases as humidity increases, due to latent
heat released as water vapor condenses when air temperatures are above
III. Hydrologic Area and Data Sources
The National Weather Service (NWS) North Central River Forecast Center
(NCRFC) is responsible for hydrologic forecasting for rivers in the Upper
Midwest and parts of the Northern Great Plains. NWS hydrologic models
and NCRFC calibrated snow and soil moisture/runoff model parameters are
used in forecasting snowmelt runoff flow into the rivers, lakes, and
reservoirs (Neuman, 1999).
River stations selected for this study, which are within the headwaters
of three of the major basins of North America, include:
Red River at Fargo, ND, headwaters to Hudson Bay
St. Louis River at Scanlon, MN, headwaters to Lake Superior
St. Croix R. at St. Croix Falls, WI, headwaters to Mississippi River
The river stations were chosen based on:
1) quality flow data from the early 1900s to current;
2) annual snowmelt runoff nearly every year;
3) location within the Upper Midwest and Northern Great Plains; and
4) author's experience & expertise gained working in hydrology in
Midwest and Great Plains.
Hydrologic characteristics of the river basins, terminology, and study
methodology are outlined in Table 1 (work sheet for Figure 1, discussed
Source of mean daily flow data was the United States Geological Survey
(USGS). Source of dewpoint data was the Midwest Regional Climate Center
IV. Timing of Annual Snowmelt Runoff
Mean daily flows were used in this study to determine �Annual Beginning
Day of Snowmelt Runoff� for years from 1910 through 2003 at the three
river stations. The methodology is explained in Table 1 (work sheet for
Figure 1 shows 10 year moving averages for annual beginning day of
snowmelt runoff at the river stations. The 10 year moving averages for
Julian Days (each Julian Day representing the beginning date of snowmelt
runoff for a year at a river station) are plotted on the 10th year of the
10 year moving Julian Day averages.
The data on Figure 1 show trends for recent earlier in the year annual
snowmelt runoff at the river stations, that began during the 1960-1980
period, and became more evident during 1981-2002 period.
V. Average Monthly Dewpoint
Climate stations that are within or near the three river basins include:
Fargo, North Dakota (within Red River basin)
Duluth, Minnesota (southeast of St. Louis River basin)
Eau Claire, Wisconsin (southeast of St. Croix River basin)
Monthly January to April dewpoint (10 year moving averages), based on
1948-2003 monthly averages, are shown in figures 2-4. The figures show
recent increasing dewpoint trends for January, February, and March 10
year moving averages, but no trends for April monthly dewpoints.
VI. Conclusions on the Timing of Snowmelt Runoff and Humidity
1) Trends were shown for recent earlier in the year annual snowmelt
runoff at three river stations within the Northern Great Plains and Upper
2) Trends were shown for recent increasing dewpoint averages for January,
February, and March but not April.
Other factors besides humidity are important in affecting snowmelt,
including air temperatures, wind speeds, temperature of precipitation,
ground temperatures, extent of snowpack over the entire Great Plains and
Midwest and its albedo (characteristics of the snow cover in reflecting
VII. Air Temperatures
Based on snowmelt physics, historical modeling, and real time operations
involving snowmelt and snowmelt runoff, air temperatures and humidity are
likely the most significant factors affecting the rate of snowmelt.
Thus some investigation and reporting on air temperatures is warranted
with respect to snowmelt. �The largest increases in both temperatures
and humidity for the Northeast, Midwest, and Northern Great Plains have
been during Winter and early Spring months� (Neuman, 2003). The report
by Neuman (2003) included selection of temperature stations and analysis
and summaries of mean air temperature and dewpoint data for many stations
in the Midwest and Northern Plains. In an investigation and report on
the climate in the Great Lakes region, from a study that was entirely
independent from the work and report on the Northeast, Midwest, and
Northern Great Plains by Neuman (2003), the Kling (2003) concluded for
the Great Lakes region that:
�In the past four years, ..., annual average temperatures have ranged
from 2 to 4� F (1 to 2� C ) warmer than the long-term average and up to 7
�F (4� C) above average in winter.�
The conclusions on temperatures by Kling (2003) and Neuman (2003) were in
agreement, even though the work was done independently.
VIII. Additional Discussion
From the snowmelt physics discussion in Section II, it is clear that
humidity and the rate of snowmelt are connected, with increases in
humidity resulting in additional heat transfer from the latent heat of
condensation as water vapor condenses on a snowpack of 0 degrees Celsius,
when air temperatures are above freezing. The process can be shown
theoretically but would require considerable work to demonstrate
experimentally or with operational hydrologic and meteorological data.
This work has shown the trend for earlier snowmelt runoff in recent
years, and the trends for higher dewpoints in recent years, but this work
has not proven that the higher average dewpoints have caused the earlier
in the year recent annual snowmelt runoffs.
Mean daily flow records that were used in the evaluation of the timing of
snowmelt runoff for the river stations in this study range from 1902 to
current, a period of 105 years of record. However, monthly dewpoint data
for this study was only available from 1948 to current, only 55 years of
record. Although figures 2-4 indicate trends for recent increasing
monthly dewpoint averages for January, February, and March, the 55 years
of record may be insufficient for making firm statements regarding long
term trends in average dewpoints.
However the river flow data records, with 105 years of record, show the
timing of annual snowmelt runoff for years that preceded the dewpoint
records used in this study. In other words, the river flow data used in
the evaluation of the timing of annual snowmelt runoff for 1902 through
1947, infer the effects of temperatures and humidity for the period 1902
In viewing Figure 1, the 1920s to early 1950s period had earlier annual
snowmelt runoffs than the late 1950s and 1960s period. However, the
period from the mid 1980s to the snowmelt runoff period in 2003 had the
earliest annual snowmelt of record, substantially earlier than the 1920s
to early 1950s period. An evaluation of mean annual dewpoints at
Minneapolis, MN from1918 to the 1940s, which were calculated from mean
annual relative humidity and annual air temperature data (Table 2.) shows
that 5 year annual dewpoint averages for 1998 through 2002 exceeded all
previous 5 year averages at Minneapolis since the beginning of record for
calculated annual dewpoint averages (1918).
The recent trends shown in this study for earlier annual snowmelt runoff
at river stations within the Upper Midwest and Northern Great Plains, and
for increasing January - April dewpoint averages call for:
1) Review of NWS hydrologic models used by NCRFC in modeling snowmelt
2) Review of NCRFC snow and soil moisture model parameters used in
models by NCRFC in issuing hydrologic forecast products with the NWS
Advanced Hydrologic Prediction System (AHPS). AHPS is described by
Deweese, M.M.. (2002) AHPS Procedures and Products at the NCRFC <
Dunne, T., Leopold, L.B. (1978) Water in Environmental Planning; pp.
Kling, G.W. et. al. (April, 2003) �Confronting Climate Change In the
Great Lakes Region <
Neuman, P.J.(1999) Hydrologic Forecast Procedures & Spring Flood
Outlooks, Upper Midwest <
Neuman P. J. (April, 2003) Special Report � Air Temperatures & Dew
Points, Great Lakes States <
Used with permission of the author.
Please send mail to webmaster@... with questions or comments
about this web site. Minnesotans For Sustainability (MFS) is not
affiliated with any government body, private, or corporate entity.
The best thing to hit the internet in years - Juno SpeedBand!
Surf the web up to FIVE TIMES FASTER!
Only $14.95/ month - visit www.juno.com to sign up today!