IAMG – spatial physical measurement subgroup

Workshop summary 

Abstract

The objectives of International Aquatic Modeling Group (IAMG) are to define the state-of-the-art in assessment and modeling methodologies for instream habitat studies.  The first activity of this kind was organized by Physical Spatial Measurement Sub-group in November of 1996 in Austria. During the week-long workshop, five teams of researchers  (IAMG-members) compared and contrasted their best available stream assessment methodologies.  Characterizing small streams and river domains was the focus of the workshop as these highly heterogeneous river types pose the greatest challenges for spatial sampling.  Two test reaches within the alpine river Ois in lower Austria were selected for methodological testing.  Quantitative results and observations from the field segment of this workshop are presented as well as technologic and methodological notes.  Additional processing of the data collected by each group will be processed in the following phase to build the background for final evaluation and conclusions. These are objectives of second workshop that will be organized in March 1997.  This very applied  results can be very valuable for the scientist working on the assessment of instream habitat. 

Introduction

Literature overview on physical habitat assessment shows relatively high inconsistency of sampling methods within scientific community. Nevertheless comparative conclusions are often built based on different data collection approaches. The inaccuracy of such comparisons that results from the above is unknown. An overview of existing technologies could help to estimate those errors and define more consistent and advanced techniques. 

The workshop was organized by the Department of Hydrobiology, Fisheries and Aquaculture (HFA) of the Universität für Bodenkultur – University of Agricultural Sciences in Vienna in Austria between 4 – 8 November 1996. 

Participants

AUSTRIA:

• Department of Hydrobiology Fisheries and Aquaculture
  • Piotr Parasiewicz
  • Christian Hofmann
  • Martin Renk
• Institute of Watermanagement, Hydrology and Hydraulic Engineering
  • Helmut Mader
  • Helmut Habersack

FRANCE:

• CEMAGREF
  • Yann Le Coarer

NORWAY:

• SINTEF NHL
  • Halvard Gundersen
  • Atle Harby

USA:

• Utah State University (USU) – Institute for Natural Systems Engineering
  • Chad P. Gubala

Narrative

During the first three days, the workshop was located in Biological Station of Austrian Academy of Sciences in Lunz am See.  During days 4 and 5 discussions were continued in the HFA department. The first day was used for theoretical presentation of methods applied by each group. On second and third day different instruments and sampling strategies were tested on the river Ois, a non-regulated, 4^th order stream. The day four and five were dedicated to discussion and organizational issues. On the last day Stefan Schmutz from HFA joined the workshop as an representative of biological sub-group.

Day 1: Theoretical presentation of methods

From the five teams that were present four methodological approaches were presented this day. As a first Parasiewicz described the methods and instrumentation that is used on the department of HFA. The irregular systematic sampling with use of cross sections have been practiced on the department since 1989. The new advance is development of DVP (Depth Velocity Position) – bar a multiplex instrument that measures integrative mean column velocity (Dipstick from Jens), near-bottom velocities (inductive-electromagnetic probe) and depth (pressure transducer) simultaneously. The data (means and time series) is recorded on datalogger. The position of the bar is recorded by total station (TS). An on-site-data-evaluation procedure that is used by HFA was presented. That AutoCAD based routine allows immediate graphic control of measured geometry.

Le Coarer presented the methods used by his team in Cemagref. The patches and their edges are sampled with TS. Sophisticated data processing, based on curvilinear location techniques of interpolation and girding, builds very accurate digital terrain model. This model offers numerous calculation possibilities and insures convivial presentations of the results. 

Harby presented methods used in SINTEF. Irregular mesohabitat sampling oriented technique is often used for calibration and validation of 2D and 3D hydraulic models, which are components of River System Simulator. Lately developed instrument Viking-Stick is used for this measurement. This multiplex instrument includes four Ott propellers for measurements  of velocity at different elevations and pressure transducer for depth measurement. After first 20 sec. of sampling the propellers are scanned in series, every four seconds one. Collected data can be either recorded on implemented data-logger or radio-transmitted to the note-book computer on the shore, where it is quality controlled. Position is recorded by TS. 

Gubala presented the latest technological advances on the field of remote sensing. Ashtech differential GPS stations offer an accuracy of about one centimeter in horizontal and about 2cm in vertical by measurement period over 40 sec. In combination with laser distance-meter (LDM) it can be a simple alternative to the total station. He addressed also limitations of GPS use.

Testing of the above become a task for the next two days. 

Day 2 Test of instruments

The objective of this day was to compare an accuracy and efficiency of available instruments for morphological and hydraulic measurements. About 200 m long site was chosen on the river Ois for comparative measurements.

Instruments

Following instruments were used this day:

0. For geometrical measurements: 

1. Total Station Leica TCA 1100 with automatic target search (ATS).

1. Differential GPS Ashtech…..

1. Laser distance meter….in combination with GPS

0. For velocity/depth measurements:

1. DVP – bar (mean column and bottom (one dimension (1D)), pressure transducer)

1. Viking Stick (four Ott propellers 1 D, pressure transducer)

1. Two dimensional probe Delft p.-e.m.s.

Georeferences:

The georeferences were defined for the site using three fixed monuments and GPS base.

Several points were benchmarked and measured with both instruments sequentially. Also simultaneous measurements of other points were done. To asses the accuracy of the instruments during collecting of hydraulic data GPS rover and TS reflector were mounted to DVP-bar and Viking-Stick respectively. After few measurements the instruments were switched between each other. 

Velocities:

About 27 verticals were measured with the Viking-Stick and DVP-bar at the same position sequentially. The measurements were stratified in three meso-habitat-type groups: pool, run, riffle. Additionally some near-bottom velocities with the two dimensional probe were collected at the same points. To asses timely flow fluctuation few five-minute time series were sampled with both instruments.

Day 3: test  OF SAMPLING Strategies

The objective of this day was to test four techniques of morphological/hydraulic data collection on the small natural river. The goal for each team was to collect the data that will give the best description of the site for the widest range of potential study objectives.

A 30m long site with three mesohabitat types (riffle, pool, run) was selected. Each group used approximately 2 hours to complete the work. Ashtech GPS and two TS Zeiss – Elta 40 (produced 1987) and Leica – TCA1100) were used simultaneously.

Following methods were tested:

1. Mader: Equidistant x-sections with 2 dimensional velocity measurements. Transect edges were sampled by means of GPS , distance between the verticals using measuring tape (10 bottom velocity points, ca. 60 depth points).

1. Parasiewicz: Irregular systematic transect sampling with DVP-bar. Position was sampled with TS. Additionally habitat parameters like substrate, cover, embededness were recorded (117 points).

1. Harby, Gundrensen: Stratified irregular systematic sampling using Viking-Stick . Position was registered with TS. Substrate was also recorded (48 points).

1. Le Correr: Irregular systematic patch sampling using  TS. (no velocity measurement) Substrate was registered. (232 points)

In addition comparative measurements with 2 D probe , DVP-bar and Viking-Stick were taken at 10 marked positions. The shore line of 200 m long reach including selected site was sampled by means of GPS and  LDM.

Day 4 & 5: Organisatory issues and conclusions

Additional test for  comparison of GPS and TS was done. Two measuring tapes of 5 and 10 m length were exposed in the garden of HFA. Several points were measured on the tapes with TS and GPS simultaneously to calculate resulting tape lengths.

Conclusions:

The first conclusions addressing sampling techniques could be defined in those days. More detailed can be set first after evaluation of collected data. 

Position:

0. Due to relatively long measurement periods TS ELTA 40 was not effective anymore.
1. TS TCA 1100 was over sophisticated and following user unfriendly.
2. Simultaneous use of one TS (TCA 1100) with two reflectors  was very efficient. 
3. Automatic target search of TS TCA 1100 was not practical in use.
4. GPS seems to be to less accurate in vertical but quite good in horizontal (measurement times between 45 – 60 seconds) – this has to be approved by collected data. 
5. GPS is more handy and saves one person (TS operator)
6. The optimum sampling team size is 

for GPS 1 specialist + 1 technician

for TS 1 specialist + 2 technicians

The minimum team size is 

for GPS 1 specialist + 1 technician

for TS 1 specialist + 1 technician

Data quality assurance is very essential and can be automated to certain level (i.e. instruments failure).

One additional person on site improves data quality.

Real time back-up system is essential and advisable.

Data radio-transmission is good solution for the back-up.

Radio communication in the field is necessary and improves data quality.

Datalogger should be simple in use not have to many buttons and options. Possibility of erasing of data on the logger should be excluded in the field.

Depth measurement  with pressure  transducer is very effective.

Velocities

DVP samples quickly many verticals with apparently lower accuracy. It measures mean column and near bottom velocity in 1D and time series at once.

Viking-Stick is more accurate, more time consuming and more sensitive to failure. It measures 4 velocities in the column in 1D at once.

2D probe has the highest accuracy, is most time consuming and problem of data storage has to be better solved. It measures  one velocity in the column at once.

2D measurements are good enough for the evaluation of 2D models and can be done well by now

Multiplex 3D probe can be easily developed, however on very high costs.

Modern hydraulic modeling requires more accurate velocity data in lower quantities combined with exact geometry

Biological modeling requires for the time being rather broad data assemblage with lower accuracy

Biological requirements on accuracy are different for preferably occupied and non occupied  locations (i.e. accuracy in boundary layer is more essential for salmonids, than in water column above).

The description of velocity habitat conditions can be done either by means of hydraulic modeling (that requires higher sampling accuracy) or by direct sampling for biological models with higher data quantity. 

Very high timely fluctuations of velocity and flow direction could be observed in the riffle areas.

Following methodological approach seems to be accurate enough for hydraulic modeling

runs and pools with 1D accuracy

riffles with 3D accuracy

From biological point of view there is no need for 3D measurement in riffles. Rather the velocity range (standard deviation and turbulence) appear more relevant. Alternative new methods like measurement of sound intensity could be effective, but need more development.

Issues to be addressed in future developments:

Instruments:

3D sensitivity

recording of time series measurements

recording of time scale in the domain

higher accuracy in occupied places

Models:

use of different models for different rivers.

Substrate:

Visual estimation is very effective approach.

Development of instruments for this purpose of substrate estimation does not appear effective. 

However there is urgent need of standardization of methods.

Development of methods for description of temporal changes is necessary.

Methodological evaluation of the objectivity of the visual estimation should be undertaken.

During discussions following questions were raised and are addressed to the other sub-groups of IAMG:

Biology:

How necessary is 3D sensitivity?

How important is velocity variation in time?

Is higher assemblage more important than higher accuracy?

Hydraulic 

What is minimal and optimal velocity measurement accuracy  for hydraulic models?

Technology transfer

Is it possible to have a universal compiler for all our different data?

Assessment:

Should we use different models for different rivers?

Temporal Changes

How many snapshots are necessary to describe the system? When they should be done?