An inspection of Gruvedalen was conducted last Thursday by Ole Humlum, an adjunct professor at The University Centre in Svalbard, along with representatives of Longyearbyen's municipal government. They wanted to see how the ground reacted to a record amount of rain that fell two days earlier. Small cracks in the tundra indicated movement, but did not indicate danger of a larger landslide. The same day, all evacuations of residences in Longyearbyen was lifted.
The following day, in his office, the professor said he and a group of geologists at UNIS expected multiple major landslides due to the weather predicted. Examining previously written material showed that in 1971 Longyearbyen was hit by a rainstorm similar to the one occurring this month. There was far less rain during the earlier storm, but there were more than 80 landslides – several of them large. In October of this year there were also several landslides considered small relatively small to the amount of precipitation.
"Now we are seeing movements, but my impression is that it was not as bad as it could have been," Humlum, said. "At the moment we're speculating what the reason is, but I have a hypothesis."
His hypothesis is frost in the top layer of the ground prevented water from getting into the soil and down to the permafrost.
"It was not much, but there was a little frost in one week," he said. "The ground was frozen, how deep we do not know."
He said he assumes the ground was frozen from five to 15 centimeters deep, depending upon the height above sea level. The mild weather and rain started to thaw the frost, but it took a few hours and only late in the storm – when the rain had subsided – did the water reach the permafrost.
It's when there is a lot of water on top of ice on top of the permafrost that friction is low and there is a significant chance of the entire active layer detaching.
"So we got lucky," Humlum said. "If we had not had frost the water would come down earlier to permafrost and we would have gotten more and probably bigger landslides. There nature helped us in a good way."
The terrain was exceptional wet for a few days.
Svalbard Airport received 46.9 millimeters of rain during 24-period from Monday, Nov. 7, to Tuesday, Nov. 8. Measurements taken by UNIS researchers in Adventdalen showed 75 millimeters of rain fell there.
"That is a third of the normal annual precipitation," Humlum said. "There is no doubt that that was an unusually high amount of water."
He said is reasonably certain UNIS' measurements are not erroneous.
"It is well known that rain gauges can produce erroneous readings when the wind is blowing," Humlum said. "But then it's just below what the results have been. Perhaps our measurements been more favorable to than the ones at the airport, but it could also be that it has rained more in the valley. As with temperature, there may be large local differences."
By comparison, 33 millimeters of rain fell during the storm in 1971.
The professor was among those scouting the hillsides in one of The Governor of Svalbard's rescue helicopters the day after the intense rainfall.
"In-between snow flurries I got the impression that the westerly terrain had received more rain than further into Adventdalen," Humlum said. "That confirms that there may be considerable local differences."
Unstable in Gruvedalen
Because of the darkness, geologists at UNIS have not yet obtained a full overview of what has happened in the hillsides around the town. What they have seen is several smaller debris slides that have stopped by themselves.
"My experience is that nine out of ten slides stop by themselves," Humlum said. "But sometimes they continue down a slope like a mud flow and can take more material with them. We cannot predict how it can happen and not."
He said that when the light comes back, it is a certainly there will be hundreds of debris settlings that become known in areas such as Gruvedalen where some homes were evacuated.
"They will show us where the ground is potentially unstable," Humlum said. "It might not happen any more after such settlings, but in places where there are buildings we have to be on the safe side."
Will the hill above the houses be more volatile in the future?
"No, now it has set itself and the water is out of the system," Humlum said. "There are small and large cracks, and when the frost comes they will help with freezing faster than it otherwise would have occurred."
The settlings are, nonetheless, signs that sloped areas may be unstable when they are saturated with large amounts of water. It may be possible to eliminate the problem in Gruvedalen with a drainage ditch up by cable car trestles.
"That can be first instituted during the summer," Humlum said. "Part of the problem is small streams from Sukkertoppen that divert water down to the buildings there. But they are such that they are difficult to see in advance. Now we are trying to understand why it happened and what we can do to lessen the problem in the future."
End of landslides at dog kennels
The most landslide damage after the rain occurred when the entire active layer behind the dog kennels just outside Longyearbyen was released. Humlum said he believes water from the drainage canal to the building at Kuhaugen brought extra amounts of water to the terrain above the kennels.
"I cannot say that the landslide would not have come anyway, but the extra water made the situation worse," Humlum said.
He said the same thing happened at the kennel as at the cemetery in October. First there was a continuous landslide. Then water entered the slide's path carrying sludge and mud that produced the subsequent debris flows.
"We can never rule anything out 100 percent, but the bulk of what might be released at the kennels has come down," Humlum said. "What we're seeing at the bottom of the landslide is permafrost. That doesn't release."
The permafrost will begin to thaw during the summer, which can trigger a little activity in the soil.
"It will take two to three years before it stabilizes, but there will be no danger," Humlum said. "Nor is it a danger that the building on Kuhaugen will be hit by a slide."
After warnings about the intense storm were issued, a large part of Lia, the area surrounding Svalbard Church, Huset and all of Nybyen were evacuated.
"We know from the past that there has been landslides in Lia," Humlum said. "In 1971 there were some landslides there. They also occurred on the mountainside opposite the old Longyearbyen. That is because the mountain slopes there are more fine-grained mass than elsewhere in the city. The drainage is inferior than coarser mass."
He said one can still see the scars from the 1971 landslides.
"But you can never say exactly where the next landslide will go," he said.
Seeking better measurements
Researchers say they see a need for more monitoring stations that provide results in real time to understand what is happening to the ground.
"In order to follow the situation closely when we get a lot of rain we have to have an online station that gives us information about rainfall and temperature in the ground from the surface down to the active layer," Humlum said.
UNIS currently has a weather station on Gruvefjellet, but it provides information with a two-hour delay. It measures the temperature from one to five meters deep into the hill, but does not indicate whether the ground is frozen or thawed between the surface and the permafrost.
"Now we are sitting a little afterwards and trying to rationalize and understand what has happened," Humlum said. "But we have a good foundation and are not engaging in guesswork."