Integrating Real-time Environmental Data into GIS-Based Military Land Management Systems.

Integrating Real-time Environmental Data into GIS-Based Military Land Management Systems.

Abstract:

On-going monitoring of Watersheds at Fort Hood Texas impacted by military training activities collects data continuously on stream level, turbidity, soil moisture, fuel stick moisture and temperature, rainfall, wind speed and director and other parameters. This data is being telemetered by cell phone to base environmental coordinators and modelers at the U.S. Army Corps of Engineers who are incorporating the data into computer models of rainfall runoff and other responses of the affected watersheds to the combination of environmental impacts and rainfall events. These models are then going to be used in helping to prioritize and assign training activities This project has involved development of GIS base maps for vegetation and soils and integration of this data along with the real-time telemetered data into models to improve decision making and stewardship so as to insure sustainable use of the unique training resources available at this base.

System Components.

The system used to gather data from watersheds at Fort Hood consists of a Forest Technology Systems remote telemetered weather station, which has been modified to accept inputs from additional stream level and turbidity and soil moisture sensors. The weather stations contain a twenty-foot mast with a mounted anemometer and wind direction sensor, humidity and temperature sensor, a solar panel and a fuel stick. A fuel stick is a ponderosa pine cylinder that contains a temperature and humidity sensors. It is designed to predict the temperature and humidity of dead wood (fuel) that may be present in a given climatic regime. The mast also supports a cell phone antenna and a solar panel. The sensors on the mast all enter a 4X4X5 foot enclosure with a watertight door secured with 16 ¾ bolts. This enclosure was custom fabricated from 3/8 steel plates and supported on legs above possible high flood levels. These enclosures were designed to be effectively bullet proof and tamper resistant and are set in concreted. There is also a tipping bucket rain gage provided by Forest technology systems mounted on the top of the enclosure. Inside the enclosure is a data logger a deep cycle battery, a voltage regulator a cell phone modem and a Motorola cell phone. The other sensors are a Growpoint sensors that measure the dielectric constant of the soil, two of which are installed at each site, a laser backscattering based turbidity sensors developed by McVan Industries in Australia mounted in concrete and set in the stream channel and bubbler based water level sensors from Hydrologic Services of Australia. These sensors use a steady stream of nitrogen bubbled out the end of a tube set at a measured depth in a steam to measure the depth of the stream. The backpressure exerted by the water above the end of this tube is measured by a pressure transducer mounted inside the enclosure. Since all these sensors are analog, an analog –digital conversion is accomplished with a custom designed device placed in a sealed box so the final output of each sensor is compliant with the SDI 12 protocol. Each sensor records changes in the measured parameter at 10-minute intervals these are then stored on the data logger and recovered daily by cell phone telemetry. Calls at user chosen times can also be made to the system. The data is being sent to both a computer at the base natural resources office and a computer located at Sam Houston State University. Once downloaded to the computer it is exported to an access format database table and made available over the Internet. Efforts to move this data to a server with automatic updates are underway. The data can then be incorporated into GIS based models and linked to point features that represent the monitoring stations. ORCALE database tables designed to store the gathered data in an appropriate schema have been are being developed as an on-going part of this effort.

Data Gathered:

Data gathered includes, stream level, turbidity, soil moisture in volumetric content over a range of 8% to 48%, rainfall, temperature, relative humidity, fuel stick temperature, fuel stick relative humidity, wind speed, and wind direction. Derived indices include hourly precipitation and 24 hour precipitation, as well as maximum and minimum 24-hour figures for all meteorological parameters.

Additional Data:

In addition to the data gathered by the remote sensors other data is being collected to calibrate and derive other indices from this primary data source. Specifically stream velocity data and stream level data based on a stage measurement are being gathered to calibrate the stream level sensors and derive ratings curves to convert stream level into stream flow at the gauging station location. This involves an accurate survey of the cross sectional area of each stream at the location of the stream level sensor and the depth of that sensor tubing outlet relative to the thalweg of the channel. Since stream dynamics can cause a variable channel cross-section and depth periodic re-survey of this channel is being undertaken. Also the soil moisture sensors require calibration this can be accomplished in several ways, sampled of soil are collected at various times and compared to the soil moisture estimates generated by the dialectic constant soil moisture probes. Also tensiometers installed near the soil moisture probes provide additional information on the moisture dynamics of these soils. GIS based soils maps derived from NRCS county level soil surveys converted into Arc/Info coverages help extrapolate soil moisture measurements at these limited locations to the watersheds as a whole. Use of the Guelph permeameter provides estimates of hydraulic conductivity at representative locations in major soil series present in each watershed. In order to calibrate the turbidity sensors as well as to be able to gather general and specific water quality parameters pumped automatic water samplers manufactured by Isco have also been installed. Thes4e samplers can take up to 24 1-liter samples in digitally labeled bottles based on a regular sampling interval. Alternatively a water level actuator switch can be installed. This approach has been adopted at fort hood so that major storm events trigger the sampler that then provides samples of water available for comparison with that being measured by the turbidity sensor. Manually collected samples collected by wading are also available for calibration as are turbidity measurements collected with a turbidity sensor manufactured by Hydrolab incorporated.

Challenges:

This project has encountered a variety of unexpected challenges. These have ranged from such minimal problems as mouse and squirrel damage and infestation of equipment (over come by better shielding of cables from sensors and blockage of entry points with steel wool) and damage to sensors from cows (overcome by attaching sensors to sturdier posts) to repeated and major vandalism. The vagaries of the weather have also proven challenging. This project had witnessed the longest and severest drought in the over 100 year history of recorded meteorological readings in central Texas a three year period of severely sub-normal rainfall that exceeds that encountered in either the Dust Bowl (centered farther North in Texas/Oklahoma) or the “The time it never rained" in the 1950’s which had its greatest impact south and west in Texas. This result of this sub normal precipitation was the complete disappearance of previously perennial streams including two of three being monitored at Fort Hood and reduction of flow in the remaining stream to a level at which the monitoring point was left high and dry. Thus stream level and turbidity data were unavailable. Major flood events also occurred during this period including one in January 2,000 that resulted in fatalities among service people within a few yards of a gauging point. Although this flood event did not damage equipment (which had been designed with just such flash floods in mind) a bulldozer slipping of the bridge over one stream being monitored did extensive damage in the spring of 2001. Specifically the analog to digital conversion device and the sensor cable and the bubbler tubing were severed and 2 inch black iron pipe used to protect the sensor cables has destroyed by the impact. Relocation of this sampling point downstream into an area better protected from traffic is underway.

Conclusion:

Despite various challenges real-time meteorological, stream level, turbidity and soil moisture data is being collected at Fort Hood, Texas and telemetered by cell phone to base computers where it can be made available to GIOS based models of erosion, rainfall, soil moisture, run-off and other parameters pertinent to prioritization and assessment of the impact of training exercises. The results of the integration of this real-time data into such models, which rely of GIS, as well as the inclusion of this data into GIS based data repositories can to help improve management of land resources on this vital military installation.

ACKNOWLEDGMENTS