C194. 1st International Conference on Urban Drainage via Internet                (c) W James, 2000.

_____________

A WEB GROUP DECISION SUPPORT SYSTEM FOR DESIGN OF BENIGN URBAN WATER SYSTEMS, USING EXPERIENCE GAINED IN AN ASYNCHRONOUS LEARNING ENVIRONMENT WITH A FOCUS ON OPTIMAL COMPLEXITY AND USER PERFORMANCE

by

William James, D.Sc, P.Eng and Eduardo Queija de Siqueira, MSc
Professor of Water Resources Engineering, and graduate student
University of Guelph, Canada N1G2W1
E: WJAMES@UOGUELPH.CA; W: WWW.EOS.UOGUELPH.CA/WEBFILES/JAMES; F: +5197672770

Last updated: 00.05.23. First published at: http://www.hydroinform.com/ICUDI/pages/papers/james.htm  Disregard the claimed copyright there, this paper is copyrighted by W James. Updated here from time to time, updates in navy blue. Latest version at my website at "Conf papers" button (scroll down to "C194"). Discussion and response appended at end.

__________________

contents [ADDED 00.05.23]

ABSTRACT
INTRODUCTION
PCSWMMGIS2000
OPTIMAL MODEL COMPLEXITY
GENERAL OUTLINE OF THE WEB GRADUATE COURSES IN JAN-MAY 2000
 Organization of the courses 
A STUDENT COMMENTARY (by Eduardo Queija de Siqueira)
USER PERFORMANCE IMPROVEMENT
CONCLUSION
REFERENCES
DISCUSSION

__________________

 

ABSTRACT

Web instructional use of a group decision support system (GDSS) for design of benign urban water systems is described. Using distributed, shared information systems on the web, the PCSWMMGIS2000 shell integrates urban water design packages in an Internet learning/design environment (ILE), and is expressly developed for improving user performance. Experience shows that achievable modelling accuracy depends more on user quality than it does on model structure. A GDSS that facilitates demonstrably improved user performance is desirable. Among other GDSS tools, sensitivity and error analysis, and on-line animated tutorials tools are included. The contextual help system exceeds 60,000 words in size. An aggregation wizard for optimising model complexity is provided.

Using this framework, nineteen graduate students from eleven countries participated during Jan-Apr 2000 in a dual graduate course delivered on one web page. The courses covered both (1) urban water pollution control planning and (2) urban stormwater management. Several pedagogic innovations were implemented and can be reviewed on-line. Several modules focussed on attributes supposed by the five instructors to improve model user performance. Links to sample student assignments and evaluations are provided. A student contribution to this paper is included.

Asynchronous ILE's are now ubiquitous. They may be designed to remove the distinctions between instruction, learning and design (work) and are an important test bed for the evolution of design technologies. An important concern here is that model user performance needs to be improved and evaluated objectively.

INTRODUCTION

The title and thrust of this paper is confusing: it covers the idea of writing models of benign urban water systems in such a way that model users perform better than they otherwise would. Ideas for the modelling approach reported here are gleaned from experience gained in web graduate courses. The paper describes certain attributes that render such a model helpful, including the idea of optimal complexity.

A group decision support system (GDSS) for the design of benign urban water systems was used in the first quarter of 2000 in a set of two web-based graduate courses. Using distributed, shared information systems on the web, the shell integrates urban water design packages in an Internet learning/design environment (ILE). The shell was expressly developed for improving user performance. If experience, which shows that achievable modelling accuracy depends more on user quality than it does on model structure, is correct, then a GDSS that facilitates demonstrably improved user performance is desirable. Such a GDSS should include tools for sensitivity and error analysis, and an aggregation wizard for optimising model complexity. The GDSS shell, (PCSWMMGIS2000), which also included on-line animated tutorials, and a contextual help system exceeding 60,000 words in size, is described in the next section.

SWMM (storm water management model) is used as the glue for the development of several new ideas presented in this paper. For the benefit of readers unfamiliar with the subject, and by way of introduction, some background information is first given on SWMM.

Development of SWMM from 1969 to date is described in Introduction to the SWMM Environment, Chap. 1 in New Techniques for Modeling the Management of Stormwater Quality Impacts (James, 1993). Development of SWMM, its ancestry and its continuing support is probably unique. Research groups at several different universities, including the University of Florida, and Oregon State University have augmented intermittent support by the USEPA (United States Environmental Protection Agency). Engineers at agencies and in consulting offices have also spasmodically contributed ideas or more materially to the evolution of SWMM. In many ways, defining SWMM has thus become like nailing the proverbial jelly to a tree. What was unusual, even at the time, was that SWMM was a major EPA funding effort devoted solely to hydrologic software development. From 1974 there was close collaboration with Canadians - notably the Ontario Ministry of the Environment, Environment Canada and several consultants and academics, also an unusual if not unique effort. However, SWMM itself remains a cumbersome accumulation of batch-oriented FORTRAN code. PCSWMM was the first user-friendly shell and personal computer version. It has been distributed with improved documentation by the writer (URL: WWW.CHI.ON.CA) since 1984, and for the Windows environment since 1993. As of Nov. 1999, it was used in about 800 design offices and 70 Universities, in about 30 countries, predominantly North America.

Probably the SWMM environment is a natural consequence of active participation in scientific, technical and engineering conferences, symposia, seminars, workshops and other meetings. Besides workshops and short courses given by (I) the University of Florida, (ii) the USEPA, and (iii) the writer, the more common meetings include irregular six-monthly user group conferences in the US and Canada. Recently, in the 1990s, meetings have been held annually in Toronto. Proceedings of these conferences are generally difficult to locate in libraries or elsewhere, so the author has made available (Biblio’98, also as part of PCSWMMGIS2000) abstracts of the papers. Papers range over many topics of interest to users of SWMM, and nicely cover the changing emphasis over the past two-and-a-half decades. Topics change from concerns with water flow quantities in a remote-batch-mainframe environment, to interdisciplinary ecosystems concerns in the still-evolving networked-workstation design environment described in this paper. One of these topics is the development of shells like PCSWMMGIS2000.

PCSWMMGIS2000

Supporting all aspects of SWMM, PCSWMMGIS2000 works in conjunction with the Rain, Temperature, Runoff, Transport, Extran, Storage/Treatment, Combine and Statistics modules. It provides a large array of file management, model development, model calibration, output interpretation/presentation, and references tools for urban drainage modelling. Some of the tools are directly relevant to the thrust of this paper and are described in detail below in this section.

File management: The file management methodology uses data distributed over local area networks (LAN's), Intranets and the Internet. It simplifies the organisation, connectivity and manipulation of the many files generated and used by SWMM. Data-aware objects represent groups of related files: the image of the object depicts the type of files it represents. These objects can be arranged logically in folders, connected to other objects, and cut/copy/paste between folders. Project information management is provided by means of note objects that can store text in ASCII, HTML or any document format. Data file structures have been simplified by eliminating the need for the Executive module command lines from all input data files. Instead, data file connectivity is displayed graphically in the PCSWMMGIS2000 window, and can be modified at any point with a simple mouse click. This eliminates the confusion and common headaches associated with misplaced interface files, wrong scratch file designations and long text-oriented commands and input.

Flexible tools registry: For PCSWMM, tools take the form of external processes that can be plugged into PCSWMMGIS2000 with the Tools Registry. This enables a completely flexible modelling environment created from third party and in-house tools, as well as the tools described below.

Plug-in tools: The sections below cover tools bundled with PCSWMMGIS2000.

Edit: The 32-bit ASCII text editor for SWMM files offers very fast load times for input or output files, of unlimited length. It also provides context-sensitive links to the SWMM online help file - pressing the F1 key in any data line invokes immediate help on the selected data line requirements. More details are provided below.

OpenURL: This tool enables access to email, web pages, ftp, and telnet from either Internet Shortcut objects, the Tools menu, or any URL address found in any object file. It can download SWMM engine updates with a mouse click, include links to related information or data sources on the user's intranet, the internet or in the user's input data files or note files, and/or provide easy modelling co-ordination with remote teams and graduate students. It requires a WWW browser and Internet access (can be dial-up).

Dynamic HGL: This tool dynamically displays the hydraulic gradeline for any sewer profile in EXTRAN output. Both dynamic simulations and static displays can be viewed at any time step. The program provides surcharge warning, and can search for surcharge events. Simulation summary tables can be viewed for the selected profile, and profiles can be easily switched or zoomed in on, even during simulation playback. Pump stations, storage facilities, weirs and orifices are supported, and property dialog boxes allow customisation of presentation and playback.

Graph: Graph creates customisable hydrographs and pollutographs directly from any SWMM binary interface file as well as from ASCII text plot files and Runoff and Extran output files. Continuous simulation output is easily and quickly analysed with unlimited zoom and pan in both the X and Y directions, and support for greater than 250,000 datapoints. Quality constituents, velocity and depth values can also be graphed, with simultaneous support for up to 6 Y-axes. The high-resolution plots can be printed, copied to the clipboard and/or saved to file in a variety of graphic and ASCII text formats.

Run SWMM: The Run SWMM tool facilitates the running of any SWMM version 4.2, 4.3, 4.4 or newer compilation. Custom, in-house versions of SWMM can also be used. Run SWMM automatically creates batch runs for multiple files and takes care of the Executive module command lines (file connectivity, module parameters and scratch file designations), reducing the chance of user errors. All aspects of the latest SWMM engine are supported, including Extran hot-start files, automatic collation of two upstream interface files, and WASP output. Command line switches provide options for the generation of plot files (ASCII text versions of interface files), as well as for invoking any PCSWMMGIS2000 plug-in tool at the conclusion of the SWMM run(s).

RainPak: RainPak software can be used to:

  1. access user-defined datalogger files (recorded by exception - no loss of detail),
  2. compute speeds and directions for individual storm cells,
  3. facilitate the analysis of the distribution of storm cell speeds,
  4. generate statistics on the computed relative frequency of storm cell direction, and
  5. generate discretized timeseries (at any time step from 1 to 60 minutes) from an array of raingauges for pasting into the Runoff module.

RainPak consists of four distinct modules: Import, Velocity, Collate and Direction. RainPak Import provides the capability for: identifying storm systems and storm cells, interpolating the various precipitation time series to a fixed, uniform time-step, and saving the interpolated precipitation time-series to a single file for use by the RainPak Velocity module. RainPak Velocity provides both visual and computational methods for determining storm cell kinematics. RainPak Collate facilitates the collation of various combinations of storm cell analysis results into a single comma-delimited file for import into spreadsheets and/or the RainPak Direction module. Finally, RainPak Direction creates rosette plots of relative frequencies of storm cell directions. Confidence in the computed storm cell velocities can be evaluated from statistical error estimates included in the software and/or by subjective analysis of the plotted vector arrays and comparison to the dynamic rain intensity bar plots generated by RainPak Velocity.

Biblio: Biblio provides instant access to reference information on 4500 papers given at SWMM users meetings and conferences. The database includes over 3500 full searchable abstracts, and two powerful search engines, and an updated user interface.

Sensitivity Wizard: This tool provides insight into the dominant processes of a particular model, and how they can change through an array of input functions (i.e. different rainfall intensities and durations). Any number of Runoff parameters can be tested automatically, reducing an unmanageable task to a few mouse clicks. Each chosen parameter is tested at five positions over user-specified ranges of uncertainty, the model runs are automatically batched and results compiled from the output files. Sensitivity results are presented in a variety of ways, including both non-linear sensitivity gradient and ranked mean sensitivity gradient plots, and can be easily printed or copied into reports at high resolution.

Calibration Wizard: The Calibration Wizard generates plots of observed against computed objective functions, complete with an array of evaluation functions, for any objective function. The program reads in a PCSWMM-generated list of observed/computed data points (extracted from SWMM computations), and the resulting plots can be printed/copied into reports.

Online help: PCSWMMGIS2000 provides a variety of information through an extensive, online, indexed help file (over 60,000 words). A series of Getting Started video tutorials (over 350MB) are included on the CD-ROM. Step-by-step instructions are available for all PCSWMMGIS2000 procedures. Known trouble spots are highlighted for quick explanations. PCSWMMGIS2000 also provides complete online help for all aspects of SWMM data file development for all modules of SWMM, as well as a tip-of-the-day feature. Up to the minute help is available through a searchable web database (over 150 modelling questions and answers) and provided through email.

GIS component:The GIS component brings both stand-alone GIS functionality and provides an adaptable link to existing GIS software and databases (e.g. ArcInfo, ArcView, MapInfo, Microstation, AutoCAD and more). Features include:

Further features not relevant to the thrust of this paper are not listed here.

OPTIMAL MODEL COMPLEXITY

Model complexity is a measure of both the number of processes activated and the number of sub-spaces (pipes and facilities) coded in the datafile. It is thus related to the size of the datafile, excluding the driving input time series (meteorology). Disaggregation is the intellectual procedure of dividing a real system into its theoretical component processes and sub-spaces. Aggregation is the reverse, the procedure of averaging a larger number of processes and sub-spaces into a smaller number of lumped sub-spaces and processes. The model is said to be less complex.

Component processes require specific input parameters, the more detailed the more the parameters required. Complexity arises when there is a large number of processes and interactions, when the relationships are non-linear, and where there are nonholonomic constraints (e.g. if a pump station in a drainage system switches on or off without regard to the system as a whole). We may think of model complexity as the product of the number of spatial sub-spaces modelled, and the number of processes modelled in each, and the number of input parameters required for each process. This is approximately and simply the total number of input parameters required in the input dataset. It does not include the input time series (e.g. rainfall or inflows) that drives the system model.

Consider a lake which has rain as input and streamflow as output, and whose water level changes from time to time. One could consider all of the Great Lakes acting as one tank (call this model GL1; here dx, dy are about 1000 km or more), or As a networked system of six lakes, linked sequentially: Lakes Superior and Michigan both flow into Lake Huron, which then flows into Lake St Clair, which then flows into Lake Erie, which then flows into Lake Ontario, which finally outflows at Kingston. - call this model GL6; now dx, dy are now about 100 km. We say that GL6 is six times more complex than GL1. Now consider modelling several thousands of the pipes in your city (for me it's Guelph) alone (call this G1000; dx will now be about 100 m). Such a model would be 103 times or more complex than GL1.

The higher the spatial resolution, the faster the frequency of the processes involved, and the finer the time resolution (dt) should be. In the Great Lakes system, response time of outflow is only slightly variable on a monthly basis. They respond to cycles of years, e.g. of wet weather. Individual thunderstorms have no measurable impact. We could set dt = 1 month. Small gutters and pipes in Guelph respond strongly to short thunderstorms, even if they last only 10 minutes. Here dt is about 1 minute. Flow in the pipe is a higher frequency process that must be modelled in G1000,and dt = 10 s. Choosing a coarse spatial resolution controls to some extent the time step, through the speed of the processes involved. Choosing the timestep controls the processes that will be involved. E.g. little ripples on the surface of the water have a natural frequency of about 10 Hz. They would be averaged to zero if dt = 1 minute or more. A short sharp thunderstorm would have a negligible rain intensity if averaged out over a year. 

Coarse models consider only longer frequency processes such as evaporation. Models that require fewer input parameters are inherently less complex (the number of input variables is related to the complexity). A rain time series for our lake model may be 100 yr. long at 1 month timesteps, about 1.2x103 timesteps. While the system model may have six component lakes each requiring just (say) ten parameters to describe their geometry. Such a system model may be said to have a complexity of 60. It is independent of the number of time steps. One of the difficulties that arises is that of attaining a parsimonious model. Recall however that the whole system is usually greater than the sum of the parts. [ADDED 00.05.23]

Once the model data is in PCSWMMGIS2000, it can be edited graphically, through the local tables, and/or with help from the Aggregation Wizard (which intelligently reduces model complexity). For reducing model complexity PCSWMMGIS can be used on its own, without an existing GIS/IMS source: new data can be entered graphically or through the local database tables, and/or imported from existing SWMM input data files and other sources.

Apart from the Aggregation wizard, which calculates equivalent parameters when aggregating two or more conduits to a single conduit, PCSWMMGIS2000 does not calculate/estimate any parameters. Also PCSWMMGIS cannot determine input parameters directly from shape files - just from the underlying attribute database. True GIS (ArcView and/or ArcInfo for example) products are best suited to this kind of analysis. Once these attributes have been determined and exist in the GIS, PCSWMMGIS2000 facilitates the data extraction and model development.

In the next section we consider the wider context of user performance improve-ment in the evolving ILE.

GENERAL OUTLINE OF THE WEB GRADUATE COURSES IN JAN-MAY 2000

In essence, the courses are very practical. Like our first course (James and Ostrowski, 1998), all evaluations are based on web pages developed and evaluated by the participants. For the modelling of urban water systems, the courses facilitate application of prior undergraduate knowledge [previously acquired in (a) fluid mechanics and hydraulics, (b) hydrology, (c) water quality, (d) urban water systems and (e) computing]. The object is to become facile at models that analyse and plan urban water systems; in particular, those that estimate flow and pollutant quantities and optimise conveyance, storage and pollutant removal structures required for storm, sanitary and combined sewer collection systems and the management of the impacts of such systems.

The URL (must be entered fully in your browser as one line) for the courses is:

HTTP://WWW.EOS.UOGUELPH.CA/WEBFILES/JAMES/HOMEPAGE/TEACHING/661/WJ661HOMEPAGE.HTML

Both course outlines are on the web. See:

HTTP://WWW.EOS.UOGUELPH.CA/WEBFILES/JAMES/HOMEPAGE/TEACHING/661/WJ661COURSEOUTLINE.HTML

Each of the two courses required six modules (one every 2 weeks), and each module comprised:

one lecture equivalent to about 6 h of lectures;
20 h of additional individual work
10 h additional technical reading;
one assignment (8 h),
and an evaluation (2 h) of every other student’s assignment.

Modules and content for the two courses were as follows:

05-661 Urban stormwater management modelling: 

1. continuous stormwater management models and model structure (SWMM and PCSWMM);
3. GIS data management, complexity, discretization and process disaggregation (PCSWMMGIS);
5. routing in complex, looped, partially surcharged pipe/channel networks (SWMM-EXTRAN);
7. pollutant build-up, washoff and transport (SWMM-RUNOFF, -TRANS);
9. pollutant removal in sewer networks, storage facilities and treatment plants (DETPOND);
11. Sewer network designs for the future; technologies for wastewater in urban infrastructure.

05-662 Water pollution control planning: 

2. philosophy underlying public water pollution;
4. developing area-wide pollution control plans and sustainable use plans;
6. introduction to BMPs and the SLAMM model;
8. introduction to the WASP model;
10. Urban litter in drainage systems; 
12. examination of quantitative and non-quantitative information in the context of planning.

The following lists the left hand toolbar, developed for all web pages in the course home page system. It neatly encapsulates the main content of the two graduate courses as posted on the web:

Daily News
* Today/home
* Modules
* Dates
* Class work
* Grades
* Course.ratings

Help
* Equations
* Sample answers
* Units
* Listserver
* Constants
* Abbreviations
* Vocabulary
* Acronyms
* Models
* Applets

About....
* This site
* Course outline
* Course index
* Questions?

Who are we?
* Instructors
* Thanks
* Class list

Organization of the courses is given below as a list of links to the course material

(all links must be entered as one complete line in your browser):

26 participants came from 11 countries, and four Universities, and 5 instructors awarded graduate credit. See:

HTTP://WWW.EOS.UOGUELPH.CA/WEBFILES/JAMES/HOMEPAGE/TEACHING/661/WJ661COURSEOUTLINE.HTML#INSTRUCTORS

The authors shared instructional activities. See

HTTP://WWW.EOS.UOGUELPH.CA/WEBFILES/JAMES/HOMEPAGE/TEACHING/661/WJ661CREDITSW00.HTML

Both courses ran from January till the end of April 2000. Work for each module was spread over 2 weeks. All dates were strictly enforced. See:

HTTP://WWW.EOS.UOGUELPH.CA/WEBFILES/JAMES/HOMEPAGE/TEACHING/661/WJ661COURSEOUTLINE.HTML#DUE DATES

Students elected one or two courses, and their work was posted on the web. See

HTTP://WWW.EOS.UOGUELPH.CA/WEBFILES/JAMES/HOMEPAGE/TEACHING/661/WJ661CLASSWORK.HTML

The instructor gradings were posted on the web. See:

HTTP://WWW.EOS.UOGUELPH.CA/WEBFILES/JAMES/HOMEPAGE/TEACHING/661/WJ661MARKS.HTML

Students became facile on the web, Internet, and HTML. At this point we present the reaction of one student, a co-author of this paper.

A STUDENT COMMENTARY (by Eduardo Queija de Siqueira)

I am a Ph.D. candidate at the University of Guelph and I was one of the students who took the web-oriented courses on urban water pollution control planning and urban stormwater management offered by the School of Engineering during January to May 2000. In this part of the paper I outline the various educational aspects that I see in this type of course and suggest some improvements.

I would describe my experience as a student in this course as very positive. It was the first time that I took a course that is quite different from traditional graduate engineering courses. Congratulations to the instructors for being innovative. I see the course as innovative not just because it included web-oriented information but because it required from the students the development of skills such as writing web-pages, web-searches, and investigating local realities, all of which are extremely useful skills for modern professionals and researchers. For most of the students in the course, I believe, it was the first time that they developed their own web pages. The necessity we had of working with html editors and managing the files in order to post them on the web developed a high level of confidence in dealing with web-oriented information. Aspects such as the importance of stand-alone web pages, and the last updated date, were typical lessons from this course.

Taking this course were students from all over the world, for example Japan, Brazil, China, Pakistan, Germany, South Africa and North America. It was interesting to note that the problems related to environmental science are so similar between different countries while at the same time having so many characteristics that are particular to each country. It was quite interesting to learn the history of sewer systems in Japan and Pakistan, and useful to become familiar with urban water systems in Southern Ontario, where I have been living for the last nine months.

I was impressed by the similarities of the educational level in the various students taking the class. We had English-speaking students but most, I guess, where students whose mother tongues were other than English. For the English-speaking participants I think that sometimes the assignments should have seemed funny, with "mislabelling". Despite that, language was definitely not a great obstacle for deriving good benefits from the course. For those who do not have English as a main language, the course was a good medium to improve English writing skills and knowledge of technical vocabulary in the field of water resources engineering. Students taking both courses needed to write one assignment as well as one evaluation of all other students' assignments per week. This constant workload including readings and writing helps in rapidly developing the required skills. I am impressed with the amount of material that we produced by the end of the course: twelve assignments and twelve sets of evaluations.

The assignments included the familiarisation with modelling tools such as PCSWMMGIS2000 (described above), SLAMM (Source Loading and Management Model), WASP (a transient biochemical model for receiving waters) and DETPOND (a detailed model for the design of detention ponds, good on pollutant removal). In a short time we were able to download and run the programs and conduct some types of analysis which students and professors see as very productive. The fact that we had five instructors in the course who were experts in the models may be one of the reasons for our rapid learning rates.

All the students in the course had to subscribe to the class listserver. For those who are not familiar with listservers, whenever one e-mails the listserver the message is forwarded to the whole class. Thus questions could be answered to the whole class at the same time, optimising use of time. Regarding questions on listservers I think that while it may work for some students it will not suit others, as some topics are not of interest of the whole class and do not need to be sent to the listserver. Students have different profiles and those differences should be preserved. Some students do not feel comfortable sending questions to the whole class and prefer sending to the instructor directly. I think this is reasonable and in this case, if the instructor thinks the issue may be of interest of the whole class, than the instructor can forward it to the listserver.

The content of the course was not much different from other graduate courses, however the fact that all the material was available on the web and can be easily accessed at any place and time is excellent. In contrast I personally, and other students too I guess, do not like long readings on the computer screen and I prefer printing the course material. I must confess it may be quite expensive for a student to print all modules based on the fact that some module printouts had more than 70 pages.

I believe that in web-courses assignments should necessarily require careful reading of the material provided. In some modules students could complete the assignment without reading the text provided.

In some assignments students were required to write technical essays about their home areas. However it is easier for German students living in Germany to obtain material for the assignment than for a Pakistani living in Canada, and these differences should be taken in to consideration. I hope that in a couple of years with the expansion of material on the web, this kind of obstacle can be managed.

Evaluation of other student's assignments can be very interesting because you learn about other realities but on a weekly basis when you have to evaluate more than six assignments on the same topic, the activity can be exhausting and not necessarily productive. My suggestion here is to reduce the number of evaluations to a reasonable number - the evaluations should simply highlight the differences between the various student efforts, or report their outstanding aspects.

USER PERFORMANCE IMPROVEMENT

Students were required to download the documentation and run the executables for WASP, DETPOND, SLAMM and PCSWMMGIS2000. Those that did experienced difficulties with the models, decreasing in order listed above. Reasons for their difficulties are clearly complex, and are conjectured below.

For the purposes of this paper module M3, the PCSWMMGIS module, is germane, and its contents are [added 00.05.17]:

Introduction
Engineering design and the development of deterministic models
Newtonian mechanics, pde's, fde's and computer programs
Spatial discretization
Temporal resolution
Averaging high-frequency processes
Model complexity
Data collection
GIS
Some urban water systems GIS concepts and terminology
Frequently asked questions
Assignment A3

The URL is:

HTTP://WWW.EOS.UOGUELPH.CA/WEBFILES/JAMES/HOMEPAGE/TEACHING/661/WJ661M3W00.HTML

The assignment of interest is A3 which was as follows:

Module M3 merges (a) modelling, (b) GIS, and (c) urban drainage infrastructure. You are required to post a web page that summarises your consideration of model complexity in terms of spatial and temporal discretization, and process dissaggregation, and of how these concerns are closely tied to the development of input datafiles, since spatio-graphical data is the essence of urban water systems modelling. Your web page should prove that you understand the terminology and concepts of GIS in urban storm water systems modelling. A huge amount of reference material is available on this subject on the web and in technical libraries in books, monographs, conference proceedings, and journals, but you do not need to spend much time searching there because reading this module will cover most of your needs. The following suggestions are offered:

1. Discuss model complexity and postulate different needs for models of varying complexity. What levels of complexity are relevant to typical design problems in your area? Show that varying the model complexity requires conceptualisation of the real problem at various degrees of spatial resolution. This implies activation of different physical processes, and requires input data files of varying size. Show that GIS could help in building datafiles because of its relationship to external databases, and its inbuilt utilities.

2. Briefly describe urban stormwater systems in your area and how they function. To help you, browse through the design tasks involved in the design of urban water infrastructure used in my undergraduate class by clicking here, and through my rough lists of urban water infrastructure, by clicking here. Please read through all of the lists and make sure that you know what each device is. If you are uncertain please check my vocabulary or get the relevant book from your library and read about the devices in question. After you are certain that you know what all the devices are, create your own local list. If possible by all means insert as many images of local infrastructure as you can (or borrow from my aesthetic infrastructure page).

3. You could try running some simple datafiles of varying complexity for your favourite model, and summarising the results - this task only for the way keen nerds.

An example of a completed graduate student assignment A3 on the PCSWMMGIS shell can be seen at:

HTTP://WWW.CROSSWINDS.NET/~JOHNPERDIKARIS/MODULE3.HTML

while an example of the same graduate student's evaluation of his peers' work on the same assignment can be seen at:

HTTP://WWW.CROSSWINDS.NET/~JOHNPERDIKARIS/EVAL3.HTML

Several innovations were included in the dual web graduate courses, as can be seen on the web. One was the large measure of self-evaluation done by the graduate students themselves. The scheme for self-evaluation of each of the assignments and peer evaluations in the two courses was as follows, where symbols are assigned the following meanings:

A = Great, I am really proud of my work in this assignment, I did better work than is normal for a grad course, even if I say so myself.

B = OK, I did all that was asked for, certainly good enough for graduate coursework.

C = Naah, less than good enough, I could and should have done more and better work.

D = This work was inadequate, it should really fail, for errors or not being completed.

Self-evaluation exercises carried out by the students were:

 

1. For every one of your own 6 assignments and 6 assignment evaluations in each course, complete the following information:

a. your self grade (12 symbols required per course)

b. the number of days late if any that your work was posted (12 integers required per course)

c. for each of your evaluations only, the number of your peers that you covered (6 integers required per course)

2. Name the one assignment that you did that you consider to have been your best (1 assignment number required per course)

3. Name the assignment evaluation that you did that you consider to have been your best (1 evaluation number required per course)

4. Name the peer whose assignment you consider to be the best for each assignment (6 names and assignment numbers required per course)

5. Name the peer (not instructor) whose evaluation of each of your own assignments you consider to have been fairest (6 names and evaluation numbers required per course).

All the above information has now been collated together with the instructor's gradings, but awaits careful interpretation.

From detailed students comments, but without offering any convincing factual arguments, it seems evident that the models ranked as follows: PCSWMMGIS2000 includes more utilities aimed at user performance improvement, SLAMM has very good user documentation, DETPOND is accompanied by good supporting documents, and WASP is difficult to find, download, install, build datafiles and run, due to the lack of user help. No objective evaluations of user performance were attempted, other than routine evaluations of assignments.

CONCLUSIONS

Asynchronous Internet learning environments (ILE's) are now ubiquitous. Many approaches are evident, and one that we used was a design that removed the distinctions between instruction, learning and design (work). We also focussed on model user performance improvement. Not all readers will necessarily approve this approach. Innovative methodology included:

1. use of models varying in support for user performance improvement,
2. lively use of a class listserver,
3. presentation of copious course materials on the web,
4. submission and grading of assignments openly on the web,
5. self-evaluation,
6. FTP sites for downloading models and documentation,
7. global problems and applications of models, and
8. exploitation of the web and its resources.


It is still too early to draw hard conclusions, but we find that our ILE comprised an interesting test bed for the evolution of web-based design technologies. An important concern here is that model user performance needs to be improved and evaluated objectively, and, as this paper hopefully demonstrates, means are at hand for this.

Finally, did the web graduate courses provide useful information for the evolution of model techniques for improving user performance in the design of benign urban water systems? We think so, but we do not yet know how to formulate an experiment to answer this question objectively. We apologise if the title of this paper is misleading.

REFERENCES [added 00.05.23]

James, W. 1993. "Introduction to the SWMM Environment", Chap. 1 in New Techniques for Modeling the Management of Stormwater Quality Impacts (W. James, Ed.) pp 1-28.
James, W. & Ostrowski, M.W. (1998) Lessons learned during a first international graduate course on sustainable urban systems modelling using asynchronous-learning on the web. UDM ’98, London, September 1998

DISCUSSION [added 00.05.23]

 

From: David P. Bernard [mailto:dbernard@essa.com]
Sent: 23 April, 2000 11:33 PM
To: wjames@uoguelph.ca; WATER-ED@LISTSERV.UOGUELPH.CA
Subject: RE: co-operative global web grad courses - delete multipostings

William, Impressive site!
We are just in the midst of a Program Review and have been experimenting with a more modular approach; clearly you have some experience in this pedagogical method.
Several questions:
1) Do you have any guidelines / heuristics for designing each module (i.e. is a module the work of 1 teaching week, 1 class...)?
2) How do students respond to having modules offered by different instructors?
3) How well does self / peer evaluation work?
Thank you.
David P. Bernard, Faculty
Environmental Science Department
Capilano College
phone: (604) 986-1911, Local 2488
fax: (604) 983-7520
http://www3.capcollege.bc.ca/envsc

REPLY

David

Thanks for your email, and comments, especially the nice starter. Sorry about the delay - I have been away.

Re: your questions:

1. I seem to have hit a good module size/frequency at last. I had my best ever course evaluations for this dual grad course. I guess that it was experience (I tried lots of combinations over the years) and the sound decision-making inherent in collaborating with 4 other Profs.

2. I think hearing it from the established expert is hard to beat. In 6 modules probably 2, 3 or even 4 Profs. would be better than 1. The problem of course is getting your colleagues to complete their share on time. Main thing is that they should enthuse very much about the teaching task, be great with students, have attractive Internet personas.

3. I suspect that self-evaluation works better than expected with most students tending to downgrade their own work. I found it useful and closely correlated with most instructors evaluations. On the other hand I have always found peer-evaluation to be the pits. Students seem to value friendly relationships very highly, too high to be objective, and I don't blame them. Furthermore their standards are as or more variable than those of our colleagues. But even worse, every now and then there's a real problem and it can end up in our version of university litigation, with real hard feelings, accusations of racism and so on - it's not worth it. Instead I ask students to read their peers' work carefully, and just say what they liked about it, or if they noticed anything noteworthy, missing or strange - just to prove they read the work critically. I don't ask them to be critical of their peers, or to give a symbol. This part is handled with ease and fun in a small class.

Well I hope that this helps, and that you let me know how you get on. I am placing a particular pedagogic paper on the course on my web page under "Conf papers" in the second column URL: http://www.eos.uoguelph.ca/webfiles/james/

Scroll down to C194. I will place all discussion there including this one if you don't mind.

Thanks again, it was nice hearing from you.

Bill James

__________________

Jan Krejcik<J.KREJCIK@HYDROINFORM.COM wrote 

Dear Prof. James and Eduardo,
first of all: thank you for your very, very interesting contribution to
ICUDI. As far as I went through the papers already presented during the
conference I found some of them quite interesting. But I was swallowed up
reading your article and, have to say, not only by reading the article but
also attending and visiting your web-courses. You wrote that "Asynchronous
ILE's are now ubiquitous". I haven't seen many of such courses like you
prepared (certainly by spending hundreds of hours) for your students. Even I
spent only "a few" hours reading some of modules from your courses I was
very impressed of its idea and content and I'd like to rise one or two
questions and maybe to add some comments on both, paper and courses.

I didn't fully understand how comments and/or results of the work of
students were taken as an "common" input to PCSWMMGIS2000 model (or any
other ) improvement/upgrade. Or did they "just" contribute to models
ranking/evaluation?

I believe that such a web-course makes lessons, learning and design more
integrated, and hence, leading to the significant improvement of user
performance. On the other hand I don't think so, that we really need to find
a way how to answer this question "objectively". The answer is in
improvement of participant skills , in their readiness to "analyze and
solve" problems in more effective way, in their commentaries (thank you for
presenting one of them within the paper..)

I believe also, that participant learned more about net-tech etc.. being
pushed to present their contributions as web pages, using list-servers, ftp
etc. On the other hand, I understand that they did it always "off-line" -
reading lessons, downloading models, using them on their comps etc.. Did you
think about "on-line DSS" having one computational/engineering example
accessible via web and launch a team-work or even some kind of
"e-brainstorming" to discuss a potential solutions and finally to agree on
"optimal" one?

Finally, thank you for exploitation of e.g. lessons about web resources and
evaluation of information that are available on the web. Some of them are
very instructive.

Have a nice day!

Ing. Jan Krejcik, Ph.D.
***************************
Hydroinform a.s.
Na vrsich 5
CZ-100 00 Prague 10
***************************
phone: +420 2 71737320
fax: +420 2 71736912
mobile: +420 603 489012
<http://www.hydroinform.com/
***************************

REPLY

Hi Jan

Thanks for reading our paper, and especially for your extraordinarily kind comments. They certainly help make my day a nice one!

Your first question, a very good one, related to how the students comments were used to improve the model. This is a part of the paper that I could not substantiate. One way It works is like this: we deal mostly in email and all serious support must be in that medium. The emails and their responses are archived (I currently have 794 items in that folder for the two grad courses) then carefully catalogued and indexed and later form part of our extensive Q&A subsystem. We do this religiously because we are very limited on the ground and have to ensure that future questions are not answered again ab initio. Another way it works is that I have a great programmer who usually immediately implements the improvements I need. Usually my request follows an instructional session with one or more students. As you can see, PCSWMMGIS2000 was written from this very philosophy - it exposes uncertainty, whereas commercial packages hide uncertainties. We hate with a passion any intellectual fuzziness, which is the antithesis of marketing.

The answer to your accompanying question about whether students "just contribute to evaluating models" is affirmative, but of course they supply copious comments on their reactions to the models too.

Finally thanks also for your question about group participation in a modelling exercise. Of course, PCSWMMGIS2000 was written for group participation through the web, but my post graduate students did not use the capability so far as I know. My undergraduates have used it, I believe. The truth may be that the design world is only waking to this possibility now, after students have used it, quite extensively in some games.

I hope that you find my answers are direct and helpful, if not please email me again. I am really pleased that you read and enjoyed our paper.

I wish you a great day too.

Sincerely,

Bill James

__________________

Standa Vanecek<S.VANECEK@HYDROINFORM.COM wrote .

Dear Prof. James and Eduardo,
the paper, you presented is very interesting. I have some question, mainly
connected to the organising issues:

1/ How the whole process of the selection of the students were organised?
Based on the number of the university, it looks like, then the students were
selected in the universities, involved somehow in the project.
2/ How the whole project is financed?
3/ Did you see any organising/financing model opening possibility for anyone
came and take fully active participation in the course? / on the other
hand - "only" reading of free published lectures and following the web-pages
created by students is externally interesting and bring loot of new
informations/
4/ Students are working on mathematical models during the course. The idea
is, then they download the models for this. This can create the problems for
using of commercial packages. Not all packages offered the demo releases
suitable for some systematic work. Some of the packages need for the proper
functionality some other softwares (GIS, AutoCAD, database.). The
alternative to this can be the modelling engine, placed on the main server
(students will send demands and obtain back result)
5/ If the mathematical model will run on the main server - it will open the
posibility for the group work - more participant working on the same
project.


Have a nice day

Standa Vanecek
Software Development and Information Systems
Hydroinform a.s.
Na vrsich 5, Prague 10
Strasnice
100 00 Czech Republic

tel 420 2 71 73 5387 ext 143
fax 420 2 71 73 4802 ext 143
mobile 420 603 439 791
email s.vanecek@hydroinform.com <mailto:s.vanecek@hydroinform.com
or s.vanecek@hif.cz <mailto:s.vanecek@hif.cz
<http://WWW.hydroinform.com

REPLY

Hi Standa

I am having a great day thanks.

Thanks also for reading my paper and especially for your interest. I will answer your 5 points seriatum:

1. Students enroll in the appropriate shared grad course at their own University under the advice of the participating professor (Bob Pitt, Manfred Ostrowski, Neil Armitage or me). We do not as a group select students, nor do we have any pre-selection criteria. We notice that quite a lot of students do not complete a course, so we believe that the course sets its own rather high standards. At Guelph I would allow any enrolled grad student to participate, but not to audit the course, and with the warning that it is pretty tough.

I invited many professors around the world to participate with us. As you see only a handful did, and they are folks who I know very well.

2. Other than the fact that we all get paid for lecturing, no funding is involved. It is done by all five of us instructors purely for the love of learning (sounds old fashioned, doesn't it?). I pay personally for all the required computer support, software, whatever. The other professors send me their copyrighted material without cost. each professor retains copyright of course, just as I do of what I am writing now. The students just pay their own graduate course registration fees at their home Universities, usually not very much.

3. There should be no problem for any grad students around the world to participate. If the grad students' University is part of the Europe-wide Erasmus arrangement, then that student I believe can register through the Technical University of Darmstadt for no additional fee, and without any prior approval. If not, the student can apply to the University of Guelph to take the course remotely. This likely will cost money and may involve some degree of bureaucracy. A better arrangement is for the grad students' advisor to email to me a request to collaborate and then we would figure out the new shared instructional load. Note that, apart from the live lectures on the home campus, most of the real class-oriented personal instruction takes place through the listserver.

4. We absolutely eschew commercial software packages, we use public-domain packages and do not know why anyone would promote commercial packages for student use. PCSWMMGIS2000 we make available to students (in any case it is extremely cheap) and it does much more than any commercial package that we know. With it, no further GIS software is required (it includes a fully-fledged standalone GIS). Remote engines are too slow and trivial for graduate work, I believe.

5. PCSWMMGIS2000 precisely allows group work through the local web browser (no grad students used it that way, however my undergraduates do).

I hope that these answers were directly useful and to the point. If not please ask again!

Well you helped make my day so I wish you a good one too.

Sincerely,

Bill James

__________________

Milan Suchanek<M.SUCHANEK@HYDROINFORM.COM wrote 

Dear William James,
your paper "A WEB GROUP DECISION SUPPORT SYSTEM FOR DESIGN OF BENIGN URBAN
WATER SYSTEMS, USING EXPERIENCE GAINED IN AN ASYNCHRONOUS LEARNING
ENVIRONMENT WITH A FOCUS ON OPTIMAL COMPLEXITY AND USER PERFORMANCE" is one
of the most interesting from my point of view. I know so few peoples from
urban waters area that have experience in using Internet in such a way. Do
you know or does anybody else know other colleagues who could share their
experience with us?

Best regards

Milan Suchanek

REPLY


Hi Milan

Thanks for reading our abstract, and especially for your kind comments.

Thanks also for your question seeking further names in the urban water sub-discipline. Sorry, I do not have any other names and addresses of folks doing this kind of work in our narrow sub-discipline. Of course in other professional disciplines they are too numerous to count.

I hope we will soon see more, however.

Sincerely,

Bill James

__________________

Milan Suchanek<M.SUCHANEK@HYDROINFORM.COM wrote 

Dear James
I am interested on your experience in WEB courses. I visited your web pages
and I was really surprised of the high level of the course. It is a good
idea to organize web courses such this was.
I would like to ask you if plan to organize next courses in the future?
Milan Suchanek

REPLY

Hi Milan

Thanks for reading my paper, and especially for your kind comments.

Yes, we certainly intend to do the course again regularly, and to make it much better.

We were very gratified with the reaction of participating graduate students, especially those from outside North America.

Thanks again for reading the course web pages.

Students are great, are they not?

Sincerely,

Bill James

__________________