Published Newsletter Index
Contents
Introduction
ADMS Papers
ADMS Validation Report
ADMS 2 Upgrade
Other Developments
Helpdesk Calls
Code Changes
Error Messages to the Screen
Surfer Bugs and Questions
User Suggestions
User Questions
Summary
This log gives details of recently presented ADMS papers, model upgrades, a record of bugs and when they were fixed and also a log of user enquiries and suggestions received over the last quarter year. This is the first quarterly log, the next will be distributed in July 1996.
"Validation of the ADMS Dispersion Model and Assessment of its Performance Relative to R-91 and ISC using Archived LIDAR data"
D.J. Carruthers et al (1995).
"Investigating the importance of pre-processing in estimating dispersion climatology"
B. M. Davies, D. J. Thomson
"ADMS-Urban: an air quality management system for traffic, domestic and industrial pollution"
C.A. McHugh, D.J. Carruthers & H.A. Edmunds.
Copies of the above were distributed with this newsletter.
| Description of Bug | Date Logged | Date Fixed/ Due to be Fixed |
|---|---|---|
| Long term average with line, area & volume sources was giving incorrect answers | ||
| Radioactivity & Particulates did not always use correct deposition parameters | ||
| Radioactivity & Specified points giving incorrect answer | ||
| Cannot Edit legend text in X-Y Plotting Facility | ||
| X-Y Plotting & Buildings - the last point on the plot is always zero, however the correct answer is written to the .C01 file | ||
| A consistent difference was found between the output calculated with the wind sector sizes 10 & 30 | ||
| Deposition was not being correctly calculated for very large particle sizes > 1000 microns. The output concentration was. | ||
| Unable to run groups and puffs together - this option was not made available initially but will be included in v2.1 | ||
| OS to terrain file was not making complete files if the grid size was 64 x 64, or merging tiles from different 20km tiles eg SS & ST |
[1] "F6413:OPEN (METDATA.OUT) - file already connected to a different unit."
This is a common error and can arise for of a number of reasons:
(a) The user has entered a pathname that doesn't exist on their hard disk or the filename is wrong or spelled incorrectly in the met. input screen.
(b) A metfile made by the user may state at the start that each line will contain more meteorological parameters than the number actually listed
(c) The met. data may be statistical when you have said it is sequential.
[2] - "OUT OF STACK SPACE"
The filename threed.vbx in the ADMS 2 directory is a different version number to the one in c:\windows\system.
Solution: Copy the most recent version number to the directory c:\windows system and remove the version from the ADMS 2 directory as it is not required to be there.
This problem has not arisen since the introduction of the install program.
[3] - "MSLANGLOAD - CANNOT FIND FILE ADMS2NH.EXE, CANNOT FIND FILE ADMS2NH.DLL"
Solution: This is sometimes due to a memory shortage because of other systems running on the computer at the same time. It has also occurred with early versions of the code which were compiled with an MSLANGLOAD statement included. It was not required to be included therefore the code was recompiled without the statement included and the problem disappeared.
[4] - "STREAMLINE TOO CLOSE SOURCE "
Occurred when using the OS to Terrain Utility and coordinates in National Grid Coordinates.
This means the source position is either outside the specified terrain file or right on the edge.
Solution: The source coordinates in the Source screen must also be in National Grid coordinates and the output grid in the Grids screen must have initial points for X and Y in National Grid Coordinates. The user will only obtain output for output points which lie within the boundaries of the terrain file. A graphical explanation of this to be included in the upgrade manual for ADMS 2.1
[5] "ERROR PLOTTING CONTOURS e.g. contours too flat to plot or stopped by user"
There may be no calculated values to plot. This can happen if the wind direction specified in the met. file does not produce contours in the user defined grid or if the run is stopped by the user. The message can also occur on machines with 8MB RAM and is due to a memory shortage preventing the launch of Surfer to produce the contour plots.
Solutions:
(a) Check that the wind direction is the one you want and then check that the output grid you have specified is correct.
(b) If the message occurs because of memory, close down all other applications, including ADMS 2 once the run is complete and launch contours.exe from file manager, or , when ADMS 2.1 has been released, from the contour plotting icon.
[1] - "APPLICATION ERROR - UNKNOWN DYNALINK"
Surfer 6 did not install itself properly. The "Restart Windows" button at the end of the Setup procedure exited from windows but did not restart it.
Solution: Exit Setup and restart windows manually - DO NOT USE CTRL+ALT+DEL at any stage during this process or the same error will happen again. A conflict arose because Surfer 6 loads later versions of *.dll files into the c:\windows\system directory than those that are in the ADMS 2 directory c:\adms-2\adms200. Users who have installed Surfer 6 should delete the following 5 ole*.dll files in the above ADMS 2 directory.
This problem has not occurred since the introduction of the install program
[2] - SURFER
When Surfer is first called it creates a blank plot named Plot 1 and then enters Plot 2.
Solution: This is a bug in Surfer and does not cause the user any problems. It will not be changed until Golden Software correct the bug.
[3] - SURFER DEFAULT CONTOURS
One customer suggested that different default values of contours could be chosen, whilst acknowledging that the user may change the contour levels once they are plotted.
Response: We feel that in the majority of cases, the default contours are sensible values. However, to choose specific values the user can select the plot by double-clicking on it and choose Level to change the contour values.
[4] - PLOTTING PERCENTILES IN SURFER
ADMS 2 allows the user to select a number of percentiles to include in the run but only one is plotted in Surfer without any indication of which one it is. How can the plot be identified and other values of percentiles be plotted?
Solution: When more than one percentile value is chosen, the contour plot produced by Surfer is the first one shown in the *.P01 file, i.e. the highest percentile value so if the user had chosen to calculate the 88th and 98th percentile, Surfer would plot the 98th percentile. In order to plot another percentile, a few changes are made to the output file within Surfer. These are:-
(i) Check the Percentiles box in the Contour Plotting screen and plot the first percentile.
(ii) In Surfer choose Grid, Data.
(iii) Delete the filenames highlighted in dark blue and type your particular filename.P01 in the appropriate directory.
(iv) Click OK and select ASCII DATA from the selection available. Click OK.
(v) In "Data Columns" under "Z:", choose D from the list. Click OK. This means that the contour plot will be of the percentile in the next column of the *.P01 file. E.g., if in ADMS you choose to calculate the 98th, 88th and 80th percentiles, the default contour plot will be of the 98th percentile. If you select D for the Z: entry, the plot will be of the 88th percentile. If you select E, then the plot will be of the 80th percentile and so on.
(vi) Three "bleeps" will be heard when Surfer has completed the new plot. To view it, select File, New, Plot.
(vii) Select Map, Contour, OK, OK and the chosen percentile is plotted. It does not have any heading however but text is easily added by the user in Surfer.
The only alternative to this at present is to do separate runs with one percentile in each. However including a list of all the percentiles available to plot, similar to the timesteps available when plotting time dependent puffs in the Contours screen is being considered for future versions, but will not be available in ADMS 2.1
[5] HOW CAN ONE OBTAIN SURFER CONTOUR PLOTS FOR Z LEVELS OTHER THAN 0m i.e. GROUND LEVEL?
Response:
(i) - In the Grids screen choose Cartesian grid.
(ii) - In number of grid points choose your own X and Y values but make the Z value = 2 (This is the maximum at present).
(iii) - In grid spacing choose your own X and Y values but make the Z spacing the height you want to plot e.g. 20m.
(iv) - Save the problem and run. Your output files will contain concentration measurements at 0m and 20m (or chosen value).
(v) - Open the .C01 file in Notepad or similar editor and delete all the headers of the file and the lines where Z = 0m. The file will now look:
| X1 | Y1 | 20 | conc1 | |
| X2 | Y2 | 20 | conc2 | etc. |
(vi) - Launch Surfer and click on Grid, Data. Delete what is highlighted in blue and type in your "filename.C01" in the appropriate directory. Click OK
(vii) - A message saying "File type not recognised" will appear. Choose ASCII from the drop down menu and click OK.
(viii) - In the next screen, click on the arrow to show the drop down menu for Z: - change it to say Column D (which is your concentration results). Click on OK.
(ix) - When Surfer has created the new grid you will hear a "bleep".
(x) -To plot the contour plot of concentration at your chosen height e.g. 20m, choose Map, Contour and the contours will be plotted with default contours.
(xi) - You can add your own title using the T (text) button in Surfer.
The inclusion of a list of the available Z values is being considered for future versions but will not be available in ADMS 2.1
[1] METEOROLOGY - WIND SECTOR SIZE
It was suggested to us that the wind sector size should not be greyed out in the meteorology screen with the default size equal to 30(. It only became black once the Long Term Average option was selected in the Options screen. The problem with that was that one could forget to return to the meteorology screen and change the size to 30( which would be appropriate if using statistical met. data. The result was that the problem could be run in a different way from that intended.
Result: We thought that this suggestion was a good idea. and a change was implemented for version 2.02. The wind sector default size was changed to be 30 degrees and the details were not greyed out. The wind sector size can now be chosen whilst in the meteorology screen which is more logical and a warning message appears if the user selects a 30 degree wind sector size with hourly sequential data which is normally supplied in 10 degree sectors.
[2]- GROUPS AND LONG TERM AVERAGE.
Some users did not realise that a single source had to be run as a group in order to calculate Long Term Average, as it was not explicitly clear in the user manual.
Result: This has been noted and will be made explicitly clear in the revised user manual for version 2.1. The table of combinations of modules that may be used together will also be revised and how to define output grids will be explained with graphical illustrations. Comments about any part of the user manual that users feel is not explained clearly and suggestions for improvement will be welcomed.
[3] DEFINING GRIDS
One user enquired whether output as an X-Y Line Plot was available using specified points when running a single source problem.
Response: The usual output files, .C01, .D01 etc. are generated and can be viewed by clicking on Results, Numerical, but the line plotting output files, .!01, .~01, .I01, .J01 and .K01 are not. The plotting output from the single source is only available with gridded output, which may be regularly or variably spaced and not with specified points. This has been made explicitly clear in subsequent editions of the user manual.
[4 ] POLLUTANTS INTERFACE
[a] A number of users suggest a pull-down menu of pollutants for easy and quick setup of .apl files. It would mean that one would not need to worry about making mistakes such as typing So2, SO2 and S02 for different sources when they are meant to be exactly the same.
Response: It is intended that such a facility will be available in future releases of ADMS 2 but not version 2.1.
[b] Users entering a number of identical sources (or with just one parameter changing for each source) strongly suggest being able to cut/copy and paste details in the input tables.
Response: This is being considered for future releases of the model but will not be available in version 2.1. In the meantime, it is possible to create an .apl file with a single source, save it, and open it in Notepad or a similar editor and cut and paste the section on source details and edit the values required for each source. You must also remember to change the "number of sources" parameter from 1, to the value you require.
[5] RUN TIMES FOR SEQUENTIAL DATA
A number of queries have come in regarding the run time of a year's hourly sequential meteorological data. On a 486, DX4 100MHz machine with 16MB RAM, a problem with two stacks, each with one pollutant and no deposition or complex effects is expected to take to take about 34 hours. This is because 1 year's sequential data has a total of 8760 hourly sequential lines.
Suggestion: A much better and quicker way to calculate an annual average concentration would be to purchase 10 years of statistical data from the Met.Office which costs approximately £390. (The same price as 1 year's hourly sequential data). In this format the data is split into approximately 2000 lines and the same run as above takes around 15-20 minutes including one percentile calculation. To calculate the worst case concentrations the user could calculate the 100th percentile. The run time is significantly increased if wet deposition is calculated, or if buildings are modelled although it is still quicker than using sequential data. We are currently developing a method to speed up the calculation of long term averages from sequential met. data.
[1] NUMBER OF POLLUTANTS x SOURCES x PARTICLE SIZES
In the user guide it states that the number of sources x pollutants x particle sizes is limited to 250. One user completed a run with a total product of 360 and called to check if the calculation was successfully completed.
Response: If an .apl file with a product greater than 250 (as above) is created in an editor such as "Notepad" (see no. 4 above), and not via the interface, then it can be run provided the user opens it in the interface and runs it without trying the SaveAs option. If the Save As option is used, the interface will check the apl file and give the following message:
The limit was set to keep down the size of runs and arrays within the code. However, apl files edited or created in an editor can successfully run with a product of 5000 (50 sources x 10 pollutants x 10 particle sizes).
[2] HOW IS THE 15 MINUTE WORST CASE AVERAGE CALCULATED USING ONE YEAR OF HOURLY SEQUENTIAL DATA?
Suggestions:
(I) If the user wants to use all the year's sequential data, they could do so by running a 15 minute average time in blocks of 99 lines of met. data (this is the maximum number of files allowed), to obtain an output file for each hour of met. data. The worst case values are written to the .C01 file and the log file .L01. It is possible to find the worst case value by looking through the .L01 files.
(ii) The user could run a long term average calculation using the whole year's data and a 15 minute averaging time and calculate the 100th percentile. The plot will show the worst case predicted at each grid point.
[3] HOW CAN ONE OBTAIN WORST CASE VALUE FOR EACH HOUR OF TWO WEEK'S SEQUENTIAL MET.DATA ?
Response: Split the data into blocks of 96 lines of data (the maximum number of output files that may be created is 99 but the meteorology should be in 24 hour blocks, 4 x 24 = 96). Ensure that the Met. data is hourly sequential box in the Meteorology screen is checked and that the Worst case value box is checked in the Output Screen. Run to obtain a log file *.L01. The worst case value for each hour in the user defined grid is written out in the *.L01 file.
ADMS SCREEN is a screening model which calculates worst case hourly averages, annual average and 98th percentiles from a single stack using prepared met. data sets. It is very easy to use and runs in minutes.
[4] WHY, WHEN RUNNING A PROBLEM WITH 24 LINES OF MET. DATA ARE ONLY 22 OUTPUT FILES OBTAINED?
Response: If this occurs, it is useful to rerun the problem and watch the screen messages that appear during the course of the run. If you have a very fast machine and are unable to read screen messages, click on NO when you receive the message "Run Terminated with Exit Code 0" and scroll back through all the messages that were written out. Common ones to look out for are:
If these messages appeared for, say, runs number 1 and 2 respectively, the .C01 file would correspond to the third line of met. data as the first two were missed out.
