FINAL VISIT, ALASKA LTER

It is not too much to say that the introduction and deployment of unlicensed Wireless into the Bonanza Creek LTER field research in 2000, has been such a resounding success, that, by the end of this project in November 2002, the University of Alaska has, on its own initiatives and with its own resources greatly expanded the places and varieties of projects now being supported by wireless links between field stations, the university, and the entire Internet community.

The original goal of this NSF funded project, which started a year later than the Puerto Rican and Wisconsin projects, was to attempt to establish one practical year-round link, in spite of expected harsh winter conditions, between several grouped environmental data logging stations 16 miles down the Tanana River and the premises of the data collection staff in one building of the University of Alaska complex in Fairbanks which supports Dr. Terry Chapin's field research projects. This was hoped to offset the labor intensive data collections efforts required by one full time person supported by boats and snowmobiles, permit data to be collected throughout the most severe winter weather, and lay down the basis for expansion of data collection sites elsewhere, all linked wirelessly.

The anticipated difficult environmental conditions included (1) the long, 16 mile, distance between university buildings and the closest data site - which distance would push the limits of the available no-license radios (2) the lack of any obvious line of site pathway, or usable relay points, between the sites and the University (3) the amount of vegetation and trees surrounding each data collection station (4) the very limited light which reaches the immediate vicinity of the data sites, both that screened by trees and the dark arctic region winters, which would limited solar power generation. (5) the lack of wind in the interior of Alaska, ruling out use of wind generators (6) the extremes of temperatures in the Fairbank's vicinity (which has a greater range than any other US city from 100 degrees F to -65, which could affect the reliability of radios and batteries.)

The only radio which promised the ability to overcome these obstacles, was the 902-928Mhz Freewave DRG series of radios. Its excellent radio sensitivity, reasonable cost ($1,000) for the support of a multiple-sensor data station, its operation in spectrum ranges that could better penetrate trees and vegetation, and its configuration flexibility made it the best initial choice. As it turned out it became the ONLY choice for 30 separate data collection sites.

The data being collected from most of the LTER sites include Air Temperature, Wind, Humidity, Soil Temperature, Soil Moisture, Snow depth, Radiation, and Rainfall.

 

Bonanza Experimental Forest Sites

It took several trial deployments over 18 months of a radio relay site roughly 8 miles from the University and four Campbell Scientific CR10X supported data site 8 miles further across and south of the Tanana River before a reliable link was established. GIS map analysis by University personnel familiar with the terrain, and ownership of land was needed.

Several sites were ruled out because of FAA objections, ownership objections, accessibility problems. In the end the intermediate radio relay site, using a 5db omni antenna mounted 20 feet off the ground on State property and right off the Parks Highway worked best. It communicated with three 5db omni antennas 8 miles further down the river, one of which was both a terminating 'slave' radio - connected to a data logger - and a further relay to a terminating radio and data logger which could not communicate directly back to the main relay site.

At the university the base Freewave radio was placed on the roof (5th floor, effectively) of the IRAC Geophysical Building and used a 12db Yagi to reach the relay point. At that point also, using a Campbell NL100 'serial to ethernet' converter, the data traffic that came via the serial Campbell Data Logger ports, through the radios at 9,600 baud, entered the University Internetwork. Thus all data loggers became theoretically accessible by researchers using PC208W software properly configured anywhere on the net.

Experimentation over the winter of 2000 and 2001 demonstrated that the solar panels and good marine 12 volt batteries - even when 4 were put in series - were insufficient to permit the Freewave Radios and its associated Campbell Data Loggers to remain in the fully on condition through the winter. Accordingly, the data loggers were programmed to turn the radios on for only a 5 minute period each hour, during which time the hourly data would be transmitted automatically to a Linux data base maintained by Ken Irving, staff programmer for the Water Environmental Research Center at the University.

In the end, this pattern was used in all other of the 30 data site deployments.

 

Caribou Poker Creek Research Watershed (CPCRW)

 

The PI of this wireless project and colleagues from the lower 48 successfully deployed in the summer of 2000 and 2001 and connected three data collection sites in the CPCRW areas 38 miles north east of Fairbanks. These sites were part Dr. Larry Hinzman's Water and Environmental Research Center, which works closely with Dr. Terry Chapin's Bonanza Creek LTER. This was the first installation in Alaska which was linked to the general Internet, and which three sites could be accessed by anyone using PC208W software (version 3.2) properly configured. This was made possible because the CPCR watershed research area is directly north and across a local highway from a permanent University Rocket Firing facility supported by NASA. This facility is linked, to the University at Fairbanks, by a series of Microwave towers. Since the facility is on the University Internet, with its own assigned IP addresses and a IP router, we were able to make arrangements for the 9,600 baud data traffic from the three data sites in the Watershed via four Freewave radios, to enter the Internet with an NL100 interface inside the Rocket Firing Blockhouse, and from thence go to the rest of the world.

The initial deployments in the Watershed were solar-equipped to attempt to last through the Alaskan winter. However, two of the sites - the ones down in the Poker Creek valley, failed in their power requirement after about 2 winter months. While the Campbell Data Loggers have been designed to operate with very low power draw, the Freewave's require almost 18 watt hours per day. Only the station on the top of Caribou Peak - which also was the key relay site, continued to operate through the winter of 2000.

Having proven out the concept that Freewaves linked through an interface to the Internet 5 miles from Caribou Peak, could reliable pass data continuously and be accessible, except when battery power was too low, the WERC technicians who had learned from this NSF project's experiences, and the web site http://wireless.oldcolo.com, took over, modified the network to produce data only hourly for 5 minutes, and it is expected the Caribou Poker Creek wireless network will function indefinitely. While initially the two Poker creek level data sites relayed their data via the Freewave on top of Caribou Peak - which had an unobstructed view of the Rocket Facility, later experiments proved that all three sites could reach the Blockhouse roof antenna satisfactorily.

The WERC technicians also were able to build on the Freewave network, by experimenting with a digital video camera - a SPARClite microprocessor driven Fujitsu camera. It was installed at the CRRL site in the CPCR watershed. By programming the camera to capture and transfer the images of the surrounding area only once an hour, recurring views of the watershed which is over 40 miles from the nearest observing station can be captured.

 

Seward Peninsula and North Slope

 

The most encouraging sign that the initial wireless successes at Bonanza Creek and the Caribou Peak Watershed has accomplished a 'technological transfer' from this NSF Wireless project to the University of Alaska - a major goal - and its staff and projects, came as B. Crane Johnson, a civil engineer hired by the WERC project, Ken Irving of its staff, and Douglas Kane who operates out of the Nome, Alaska University have undertaken to deploy, successfully, additional Freewave deployments in extreme northern Alaska. And further been able to link some of them through the Starband satellite Internet service. They have been able to 'push the envelope' with relayed links, some elements of which span 80 miles between radios, and extend northeastward from Nome over 200 miles. All providing hourly reliable meterological data. Finding that the $500 Campbell Scientific NL100 serial-ethernet device, which links the serial Freewaves to Ethernet Internet ports was sometimes unreliable, Johnson tracked down a simpler, and DHCP capable serial-internet converter from Moxa Technologies (www.moxa.com). Their 211 series device, which operates up to 10mbps, and costs $289, has worked reliably ever since.

These 6 wireless nets, with 30 Freewave radios supporting them include Seward Peninsula, North Slope, Teller, Kobuk, Caribou-Poker Creek, and Bonanza Creek.

Concurrent with these deployments, which would require, to be accessed, PC208W software installed on the Windows machines of all interested Researchers, Johnson and Irving, using Pearl script language, a Linux server at Fairbanks, have captured the data from all of the data loggers in 'raw' numerical form, then imported it to a Web site at the University (http://www.uaf.edu/water/projects/atlas/metdata/atlasmetsitemap.htm).

This converting of hourly captured data from Campbell specific form, to a generalized series of web pages, makes all the data available to anyone in the world with a web browser and the URL above.

As reported to me by Dr. Larry Hinzman, this has had a large civic impact, as Alaskans, in particular those living near Nome, have come to rely, using their home Internet systems on the weather data for vital travel data. Even the US Weather Service now uses the data from regions hitherto very hard to get timely data from.

It is fully intended by the WERC project to extend wireless links to all fifty remote meteorological station they service - ten on the Seward Peninsula, fifteen on the North Slope as far as Prudoe Bay, two near Ivotuk, and the remainder of the 20 near Fairbanks.

TRIP CONCLUSION

Both Dr. Terry Chapin and Dr. Larry Hinzman strongly endorse the use of wireless for their difficult-region environmental research. Now and in the future. The experiments under this NSF Wireless project for the Bonanza Creek LTER and in support of the Water and Environmental Research Center have more than justified its effort. It have proven that even under the most severe climate conditions that wireless can reliably support the field science being undertaken there.