Paper presented at International Conference and Exhibition on Performance of Ships and Structures in Ice, Banff, Canada, July 2014
A Strandberg1, R Wang2, P Spencer3, and G Strandberg4
As presented at the International Conference and Exhibition on Performance of Ships and Structures in Ice, Banff, AB, Canada, July 28-31, 2014
Changes in air temperatures from 1970 to 2013 due to climate change in the Northern regions can have a serious effect on the construction and use of ice roads. The construction of ice roads in the early winter season is dependent on the ability of the area weather to freeze a working surface for the use of heavy equipment. A safe thickness of frozen ground and/or floating ice sheet is required. Late winter season weather governs the last date that the ice road system can be used. This paper uses weather patterns over the 43 year period from 1970 to 2013 to determine trends that may be present in the geographical areas of Prudhoe Bay (USA), the Mackenzie Delta (Canada) and the Yamal Peninsula (Russia). In addition, the effect of these changes in the weather on ice construction procedures in these three geographical locations are discussed and summarized.
Ice road; Prudhoe Bay; Mackenzie Delta; Yamal Peninsula; Climate change.
In this paper three different Arctic geographic areas are examined for ice road construction and operation in relation to climate change from the 1970/80s to 2013. These areas are as follows (illustrated in Figure 1).
The location of the Prudhoe Bay tundra roads with zones of operation are shown in Figure 1 and Figure 2. The zones of tundra road operation for Prudhoe Bay vary in terms of elevation and coastal location. As seen in Figure 2 there are 4 zones of operation. This paper will focus on the eastern coastal zones (Latitude 70.0° N and longitude 149° W). The coastal zones require a snow depth of 6 inches and a soil temperature of -5°C for allowed operation while the foothills areas require -5°C soil temperature with 9 inches of snow depth for operation.
The Inuvik to Tuktoyaktuk ice road is a 187 km long ice road floating on fresh water on the east arm of the Mackenzie River and then onto ocean shoreline ice near Tuktoyaktuk (Tuk). The location of this ice road is shown in Figure 1 and Figure 3. Tuk is at Latitude 69°26ʹ N and longitude 133°01ʹ W. Also in this area the Peel River and Mackenzie River ice bridges are examined in terms of climate change and operation. These two ice bridges are located on the Dempster highway approximately 150 km SW of Inuvik.
The location of the Yamal Peninsula is shown in Figure 1. This area is selected to examine the effects of climate change on the other side of the Arctic from Prudhoe Bay and Inuvik/Tuk. The Yamal Peninsula weather station is Im.M.V.Popova station (WMO index 20667) located at Latitude 73°2ʹ E and Longitude 70° N. Average monthly temperatures from the 1970s to present were calculated to examine the effect of global warming. Conclusions are drawn on trends for cold weather operation from this weather data.
Figure 1 : Locations of Prudhone Bay, Tuktoyaktuk and Yamal Peninsula
Figure 2 : Prudhoe Bay Tundra roads information for 03 Feb 2014 from the site http://www.arctic-transportation.org/map-xml.php
Figure 3 : Peel to Mackenzie bridge to Inuvik to Tuktoyaktuk ice road.
The average monthly temperature for January for Deadhorse, Alaska from 1983 to 2013 is given in Figure 4. A linear regression analysis for the 1983 to 2013 monthly average temperature data is shown on this figure. The slope of the linear regression line is the change in average monthly temperature for January from 1983 to 2013. Similar graphs were produced for each of the winter months from October to May. The slopes of the linear regression line for the eight winter season months are presented in Figure 5.
In Figure 5 it is observed that the January, February and March average monthly temperatures have become colder by 0.06°C to 0.07°C per year from 1983 to 2013. Over the 30 year period from 1983 to 2013 this amounts to a 2°C decrease in average monthly temperature for January, February and March. The shoulder winter season months, especially October and November have become warmer on monthly average by 0.16°C to 0.21°C per year. The November increase 0.21°C per year amounts to 6.3°C increase in average monthly temperature over the 30 year period.
The sustained cold air temperatures in January, February and March indicate that projects that depend on ice making during mid-winter will have similar efficiencies to 1970/80s operations.
Figure 4 : Average January temperature for Deadhorse from 1983 to 2013
Figure 5 : Changes in monthly average temprature per year for winter months from 1983 to 2013 for Deadhorse.
The Prudhoe Bay eastern coastal region is shown in Figure 2. This area contains the Deadhorse weather station.
As seen in Table 1 the eastern and western coastal regions have similar opening dates until approximately 2002. In 2005 Alaska Dept. of Natural Resources implemented stricter controls on Tundra road operation with minimum required Tundra soil temperature and required depth of snow cover. This paper examines Tundra road operation in the eastern coastal region for the period before these controls were implemented from the 1970s to 2005. Deadhorse climate data is examined from 1973 to present. Prudhoe Bay use of Tundra roads after 2005 with stricter -5°C soil temperature restriction is discussed by Lilly et al. 2008.
Present Prudhoe Bay Tundra road operation restrictions require -5°C for soil temperature for most roads however this can be changed to -2.7°C in some cases. At present the snow cover requirement does follow government requirements but variation from these requirements is possible depending on a particular situation. In terms of Tundra roads from the 1980s to 2005 which is the subject of this paper, companies that operated did so with a close eye on government restrictions which became more limiting as time progressed. In the 1970s and 1980s companies were told that even though the open period had started they were still responsible for any damage that was done. Tundra damage was occurring on some jobs and permitted companies were getting heavily fined after the season. This resulted in more caution on the government end on the opening and closing dates as the years progressed.
Table 2 presents eastern coastal area Tundra road opening and closing dates from 1970 to 2004. Figure 6 presents the data from Table 2 with a linear regression analysis. Over this 44 year period, the Tundra road opening date has advanced by 2.2 days per year and the closing date has retreated by 0.44 days per year for a total loss of operation window of 2.6 days per year. The loss of 2.6 days per year of operation window is also seen in Figure 7.
Table 1 : Prudhoe Bay coastal regions Tundra road opening and closing dates (DNR 2005).
Table 2 : Prudhoe Bay eastern coastal region Tundra road opening and closing dates.
Figure 6 : North Slope eastern coastal region Tundra roads opening and closing dates from 1970 to 2004.
Figure 7 : North Slope eastern coastal region Tundra roads operating window 1970 to 2004.
The eastern coastal area of Tundra road operation for Prudhoe Bay has been significantly impacted by a warming trend from 1970 to 2005. Linear statistical analysis of the opening and closing dates of the eastern coastal area Tundra road revealed that the operating window has reduced by 2.6 days per year during this 35 year period. This reduction is due primarily to a warming trend in the October to November time period. Stricter controls did account for part of this reduction in operating window however the steady progression in loss of operating window as seen in Figure 7 points to climate effect.
The December average temperature at Inuvik from 1973 to 2013 is given in Figure 8 with linear regression analysis. The slope of the linear regression line is 0.12 ºC per year. The same calculation was done for all months from October to May with the results given in Figure 9.
Figure 8 : Monthly average tempratures in December for Inuvik from 1973 to 2013
Figure 9 : Change in monthly average temprature per year from 1973 to 2013 for Inuvik
The Inuvik to Tuktoyaktuk ice road opening and closing dates from 1980 to 2009 are given in Figure 10 with the operation window given in Figure 11.
Figure 12 presents the FDD (Freezing Degree Days) on the date of the opening of the Inuvik/Tuk ice road. Prior to 1990 the FDD at opening was typically greater than 1500°C-days and in some cases was greater than 2000°C-days. After 1990 this ice road was opened with an FDD of less than 1300°C-days. We can see that this ice road was not opened as early as could be achieved before 1990. Only the opening dates after 1990 are significant in observing the effects of climate variation.
Figure 10 : Inuvik to Tuk ice road opening and closing dates from 1980 to 2009.
Figure 11 : Inuvik/Tuk ice road opening window from 1973 to 2013.
From Figure 10 it is observed that the Inuvik/Tuk ice road opening date has advanced (or increased the operating window) by 0.54 days per year over the period 1990 to 2009 as seen in Figure 11. Over this 19 year operation period this indicates that the ice operation window has lengthened by 10 days due to variation in opening date which is a significant amount.
The Inuvik to Tuk ice road closing date has advanced (or shorten the operation duration) by 0.11 days per year from 1980 to 2009. Over this 29 year operation period this indicates that the ice operation window has shortened by 3.2 days due to variation in closure date which is not a significant amount.
Since the Inuvik to Tuk ice road opening dates are skewed due to variation in FDD at opening prior to 1990, the opening and closing dates for the ice bridges at the Mackenzie River and Peel River on the Dempster Highway were examined. The opening and closing dates for these two ice bridges are presented in Figure 13. As can be seen the opening date for these two ice bridges tends to be approximately December 10 to 20 for 5 tonne load limit with the Mackenzie River ice bridge at approximately November 20. There was no significant variation in the opening dates of these two ice bridges from 1975 to 2010 as shown in Figure 13. Improved construction methods with better flooding and spraying pumps have probably allowed the opening date to be maintained. Two pictures of Inuvik to Tuk ice road conditions in April 2014 are shown in Figure 14 and Figure 15.
Figure 12 : FDD at opening of Inuvik Ice road.
Figure 13 : Mackenzie and Peel Ice Bridge operating dates from 1975 to 2009.
For the time period 1973 to 2013 the Inuvik area has experienced a significant warming trend for the winter season months from October to May. The months of October, November and December show a pronounced warming trend of approximately 0.11°C per year for the period 1973 to 2013 (or 4.4°C increase in average monthly temperature in 40 years).
The ice road opening dates for the Inuvik/Tuk ice road and the ice bridges at Peel River and at the Mackenzie River on the Dempster Highway have not been significantly impacted by this warming trend. Improved ice making techniques and changes in operation have compensated for the warming conditions to allow the opening dates to be maintained.
Figure 14 : Inuvik to Tuk ice road along Mackenzie River in April 2014.
Figure 15 : Inuvik to Tuk ice road near to BarC in April 2014.
The Peel and Mackenzie ice bridges are seen to essentially close on the same date (Figure 13). This closing date has not significantly varied from 1975 to 2010 which is approximately day 210 after October 1 or April 29. The closing date for the Inuvik to Tuk ice road also has not significantly varied from April 29 during the period from 1975 to 2010.
The following quote from Rawlings et al (2009) highlights the problems of assigning climate change factors to ice roads in the area from Yellowknife/ Norman Wells/ Inuvik.
"Thus the strongest evidence (Nahanni, Wha Ti, and Gameti roads) is that one of three roads was experiencing a lengthening of operating season, in the face of a general warming of the climate across the Canadian Arctic.
If one relaxes the importance of the statistical significance, there is other evidence of roads "bucking the expected trend"- 9 of 19 showed a lengthening trend, whether statistically significant or not, 2 showed no trend, and 8 showed a shortening of operating season.
These first analyses indicated that examining the data in terms of global warming, measured only by the mean annual temperature, was not sufficient to provide the full picture. While the mean annual temperatures and winter temperatures (Dec-Jan-Feb) were rising, spring temperatures (Mar-Apr-May) were actually cooling. The overall trend is a rise in mean temperature, but the cold season was "moving" in time - coming later in the calendar year.
Reflected by the changes in opening and closing dates, the operating season could be seen to be moving in time as well as lengthening or shortening. The season lengthened or shortened according to the relative movement of the start and end dates. In two of three statistically-significant cases, the operating season was shifting to later in the calendar year, while simultaneously becoming shorter."
This paper focuses on climate change of the Yamal Peninsula in terms of potential ice road operation relative to Prudhoe Bay and Inuvik. No actual Yamal ice road cases are referenced.
The January average temperature for Yamal from 1970 to 2009 is given in Figure 16 with linear regression analysis. The slope of the linear regression line is 0.13°C per year. The same calculation was done for all months from October to May with the results given in Figure 17.
Figure 16 : January average temprature for Yamal Peninsula from 1970 to 2009.
Figure 17 : Change in monthly average temprature per year from 1970 to 2009 for Yamal at Station 20667.
The air temperature data from the 20667 weather station at Yamal for1970 to 2009 shows a change that is dissimilar to the results from Prudhoe Bay. The mid-winter months of January, February and March for Yamal have pronounced warming of average monthly temperature in the order of 0.08°C to 0.13°C per year. The Yamal winter shoulder seasons in both spring and fall show a warming trend in the order of 0.04°C to 0.08°C per year.
The Yamal area as represented by the climate data from weather station 20667 has experienced a significant warming trend that gives caution to plans to operate Tundra/ice roads.
Individual conclusions on each of the three areas discussed in this paper (Yamal, Inuvik and Prudhoe Bay) in the three sections are given above. In terms of overall conclusions we can conclude that Tundra roads are more susceptible to warming trends than floating ice road. Prudhoe Bay is an example of an area that has lost Tundra road operation window due to climate change. The warming trends in the fall season have delayed the development of snow cover which has compounded the problem as the lack of fall season snow cover has increased the area surface albedo which has caused a further warming influence.
Some floating ice roads at ocean shorelines and on freshwater lakes have been able to compensate for climate warming trends with improved flooding techniques. The consistent opening and closing dates on the Peel and Mackenzie ice bridges and Inuvik/Tuk ice road from the 1970s/80s to present is an example of this fact. Other inland ice roads in the Yellowknife to Norman Wells to Inuvik area have experienced inconsistent trends which highlighted the danger of assigning climate variation as a single strict ice road operation controlling factor. Other factors such as variation in government regulations, demand for the ice or tundra road and improved ice construction techniques must also be considered.
The warming trends during the winter season are different for the three areas that have been discussed here. Thus temperature data for the particular area should be used rather than relying on arctic wide trends.