PROGRESS DIARY #10

June 19-22

RECONNAISSANCE IN ALASKA

The Biological Sciences Division of the NSF requested that I, as PI of ANI-9909218 funded by the Advanced Networking division, consider adding the Bonanza Creek, Alaska LTER, outside of Fairbanks, to my work over the next two years, supplementing the original grant to cover its costs. To attempt to connect up, wirelessly, a series of data loggers that are deployed in extreme-weather places which can be very difficult to service on a regular basis. The general location is shown below.

Between June 18th and June 23d I traveled to Alaska to see the data logger sites up close and personal, consult with the PIs and Researchers for the project, judge whether or not I can connect them up over the distances required and under the conditions the radios will experience, and make an estimate of what it would take.

Since the final decision has not been reached whether this task will be added to the Trout Lake and El Yunque Research areas, I will only give a survey here of what I found. Well illustrated by digital photographs I took.

I spent one day, the 20th of June, in Anchorage before traveling to Fairbanks, the main campus of the University of Alaska, where the LTER is headquartered, and housed in the Arctic Health Center building. The purpose was to visit possible sources of supply, other than radios, that would be needed in any extensive wireless project. Power supplies - batteries, solar panels, wind generators, RF cabling, connectors, fasteners, telescoping antenna poles, and all the small items needed, that should not have to be shipped from the Lower 48.

Alaska Battery company, and Frigid North, an electrical supply house were the best stocked, and deliver quickly to Fairbanks. And demonstrate expertise about severe weather conditions which we would need.

I also accepted a recommendation from a long standing colleague, 'Red' Boucher, owner of Alaska Wireless, who had successfully installed wireless systems in villages such as Tooksook Bay, on the Bering Sea, to consider subcontracting with a technician who lives in Anchorage. I could use his knowledge of cold weather installations, his physical help when the time comes to install radios between links miles apart, and someone to be on call for trouble shooting only 1 hour flight time from Fairbanks, rather than the lengthy trips from Colorado. So Rick Erickson accompanied me to Fairbanks, and proved valuable in my preliminary site survey. We flew low over the spectacular Denali Range enroute to Fairbanks.

During the preliminary meeting with researchers from the Institute of Arctic Biology, Dr Marilyn Walker, Dr. Larry Hinzman, (Dr Terry Chapin, the overall PI was out of Fairbanks at the time, but kept close tabs on the visit) and Stephanie Pike, Data Manager, I agreed to visit, not only the Bonanza Creek area, 20 some miles southwest of Fairbanks, but also the Caribou/Poker Flat Watershed, 35 miles northeast. They requested that I consider connecting up key points in both areas, which are administered by researchers a bit differently.

We first traveled, by truck and power boat, down the Tanana River, to reach data logger sites on both sides of the river, which gets to be a half mile wide in places. Stephanie Pike, the Data Manager for Bonanza Creek sites was our guide. She travels to these sites weekly, winter (by snowmobile and snowshoe) and summer (by power boat and hiking) to bring back the data logger modules, replace batteries, and generally maintains the systems.

Stephanie Pike on her boat.

 

 

The Tanana River

We passed several data logger locations, most of which are not very far from the river, while I looked over the terrain, the density of woods, the relationship of them to peaks nearer the University on the northwestern corner of Fairbanks, and questioned Stephanie.

We traveled 16 miles downriver in the jet powered Sprint Powerboat. Jets are required because the glacially fed river carries much debris from the banks that are constantly being eroded. Here is a look at the banks of the river, where large trees grow, and the erosion from the river cuts back until they topple.

We stopped and hiked into several sites. Two of them were identified as the highest priority to link up, because they are furthest and hardest to get to in the winter. Data Logger FP4A is shown below here.

The woods are 'medium' around some of these sites, as shown.

It appeared to me it would not be too difficult to get radio links through the woods to the river, for sites on the south side, from where clean line of sites exist to some peaks. For others it would take tall masts, at least 18 to 50 feet, to clear the trees sufficiently to reach relay peaks. This will have to be confirmed by carrying actual radios, batteries, a variety of antennas, and telescoping poles on a later trip.

Mosquitos are heavy, so we were prepared with salves, and we had to wear waterproof boots across the spongy wet ground.

At a larger LTER site on the north side of the river, an area about 100 meters in diameter had been cleared (apparently earlier, helicopters were used to lower heavier tower and small building equipment into sites.) The following Tower, with 10, 3, 1 meter wind anemometers sits in the middle.

This important site is connected by cell phone, at its maximum range - and then only by use of Yagi antennas, and repeated attempts to dial, from the University. Beyond this distance, all sites would have to be linked wirelessly through relay links. Cell phone access to the data loggers is impossible.

One curious thing I noted, was how vertical to the ground the few solar panels erected were. I am not sure how exact the solar power engineering has been, for the panels are too small to fully recharge the batteries deployed at each site, but the reason given is that in winter, the strongest sunlight is at the horizon. Their strategy seems to be to have heavy duty cycle batteries, and more than one, at each site, and replace them. Hence pointing the panels. The next picture shows Stephanie in front of one panel, and pointing at where the sun, in winter, would be.

As we returned the 16 miles upstream again it seemed to me that linking the data loggers from their wooded sites to each other through the trees as ground level would not be too difficult, but getting long links back to the University will be a challenge. There is no single high point that can 'see' all Bonanza Creek data logger locations.

In the winter, they must wait until the river freezes over hard enough, before they use it as a snowmobile highway to reach the sites. And in the spring, it takes over a month after the breakup, before they can use the boat again. Of course Stephanie Pike seems quite comfortable in the winter weather, and, if required, would use her dog sled.

After meeting in the afternoon with the Researcher again, it was clear that they desire to deploy many more data loggers. But with the labor-intensive effort required to visit, weekly, each logger, there is a point of diminishing returns. So they really could use wireless access to the data, as extensively as possible. As we planned the next day's trip to Caribou/Poker Creeks Research Area, Dr. Hinzman, PI for that research identified for us three high priority data logger sites in the area, which he wished could deliver data continuously.

Caribou Creek/Poker Creek

We left at 8:30 AM with Dr. Kenji Yoshikawa, whose field is the study of Permafrost. Kenji, a Japanese American who arrived at Pt Barrow, Alaska by sailing his own boat from Hokkaido, Japan, does much of the field work in this area. He is also, helpfully, a Radio Ham operator, who has experimented with radios to move data. I realized he will probably be the one in the LTER team who will most quickly master what it takes to expand a wireless network beyond our initial installations.

We drove over paved highway out of Fairbanks, past a point where the Alaskan Pipeline is above ground and fully visible, to an unusual facility, the Poker Flat Rocket Research Range.

This facility is owned by the University, but whose work is supported by NASA and NOAA. The facility fires rockets such as the below to probe the higher atmosphere, particularly during the Aurora Borealis phenomenon.

 

 

A picture of the Aurora from the Museum at the University.

 

 

There we left the truck, and mounted three ATVs for travel into the watershed research area.

This is the only feasible form of transportation past some low lying areas, to our ultimate destination, Caribou Peak. In the lower reaches the path looks like this, but gets very rutted and rock strewn, with downed trees higher up.

One main site on the Poker Creek side of the watershed is extensive, with a CR21X handling lots of data.

This site is in clear line of site to the top of Caribou Peak, where another data logger location is.

It took us about half an hour with these powerful small ATVs to reach the peak, where the sparsity of vegetation, and a threatening lighting and thunderstorm moving in, told me how severe the environment must be on the peak. But that also makes the radio line of sight issues a lot simpler, for a radio stand would not have to be very high. Here is the data logger, with a CR23X on Caribou Peak.

We were able to see, 4 miles away, the Rocket Research site clearly. And we could just see the first data logger site we had passed, on Poker Creek, about 3 miles away. But we could not see the second site up Little Poker Creek, although a map analysis says it should not be masked by the terrain, only, perhaps, tall trees down in the valley.

Once again, only by bringing pairs of radios, batteries and antennas, can we validate link paths.

With the weather closing in, we headed back to the Rocket facility, and met with its site managers. For we had been told that there was a full T-1 Internet connection between the Rocket Launch facility and the University - the Geophysical Institute. If we could connect from the data logger to Caribou Peak, and then from the Peak to the Blockhouse at the Rocket Facility, we could solve the huge problem of getting the data, serial radio collected, back to the Arctic Biology Center. This would take some serial, to ethernet and IP network interface work, and back again into the PCs, which we had not done before, but which would certainly be feasible at modest cost.

The Rocket Research staff expressed a willingness to cooperate fully. Here is look back at Caribou Peak from the facility.

I concluded that we could usefully connect up the data loggers identified by Drs. Hinzman and Yoshikawa adequately, perhaps more easily than those in Bonanza Creek Experimental Forest. And so returned to the University to hold our exit interview, where I laid out what it would take to do the job, which, in essence, would be in three phases over two years.

Phase I -late summer 2000 to spring 2001 - a second, more technical site survey by Hughes and Erickson, accompanied by Pike and Yoshikawa. Programming over the winter to do the interfacing which would be needed at the University to move the data into their data bases and web sites, in XML language, as it continuously comes in via wireless.

Phase II - Summer of 2001 to spring 2002. Installation of radios at both sites in the summer of 2001. Monitoring and measuring performance, completing the data links to the Internet. Observing the reliability of the radio links over the winter of 2001-2.

Phase III - Spring and summer 2002. Assisting the LTER staff in installing new radio links to new data logger sites in the early summer of 2002. Observing the full year performance of the wireless links. Refining installations for the next winter. Making the final report on the Project by August 31, 2002, the end of the project.

A seeming consensus having been reached, I returned to Colorado, and drafted the Supplemental Proposal for submission to the NSF for immediate action. But not before watching Mt. McKinley out the aircraft window on the way back.

 

PREVIOUS

NEXT