Tab::General options

Date and Index

This selection affects the available wildcards in ENV and MCM input, BG particles and Losses file. In case "Index" is used, only wildcard <i> is in use, with date you have >y>, <m> and <d> for year, month and day.

The Date or Index is used together with the Group name to form the name of the output directory which will contain the Run name directory.

The most important distinction of Date or Index comes from the way the Solar Zenith Angle (SZA) is calculated in the model

• With Date, SZA is calculated based on the date, time of day and latitude.
• With Index, SZA is always 0°, so the Sun (or more probably, the UV lamp) is shining directly from above.

SZA affects the actinic flux calculation (see also Surface albedo).

Group name

This name will appear in front of the output directory which will contain the "Run name" directory. The GUI changes all names to all caps to avoid confusion, but the configuration file can use any names (without white character)

Run name

This directory will contain the output files which the model produces. The GUI changes all names to all caps to avoid confusion, but the configuration file can use any names (without white character)

Description

Maximum of 1000 characters long description of the simulation.

Common root

The Common root is the directory which will contain the Group and Run directories. It is by default INOUT in the ARCA main directory, but it could be anywhere on your computer. If the Common out does not exist, you must create it, the GUI won't do this. This is a safety precaution.

Output file names and how the output directories are named based on the Group and Run names and Date/Index

ENV input and MCM input

Using input files is not absolute necessity if you want to assign time dependent values to input variables, but most often many of the input to the model comes from measurements. For one dimensional variables these can be sent to the model with ENV input and MCM input. The ENV input is a collection of meteorological and physical variables, and (mostly) inorganic gases.

The MCM input contains organic compounds that are used (for the most part) only in the chemistry module.

The format of the input files is described in the description of tab Time dependent input. All variables that can be loaded from the ENV input are marked with blue and MCM input with beige background on the tab Time dependent input→Variables.

Note that all variables could come from one file, in which case the same file path is given in ENV and MCM inputs.

BG Particles

If the background concentrations are to be read from a measurement (or other) file, define the file here. The structure of the file is assumed to follow the SMEAR measurement station sumfile, described in more detail here. Essentially, the file contains the size distribution as dN/dLog(Dp).

Commenting ENV and MCM Input files

The first (and only first) line can - and really should - be commented to make identification of columns easier. The comment starts with #, and each column should have one word name. Best to use space as separator. If the columns are named properly, this will help great deal to define the correct column content to correct variable using the tool Print input headers with column numbers.

Using wildcards in file names

You can insert (parts of or all) of the Date or Index using wildcards, see here for detailed explanation. Wildcards are parsed from ENV input MCM Input, BG Particles and Losses file (later in tab Losses). The figure below shows an example of wildcards in use.

Since the Date is now 2018-04-11, the BG particles filename will be parsed as "dm180411.sum, as shown by the tooltip, which shows the parsed path.

File time units

These options set the time unit in the ENV and MCM files, and the measurement interval in the DMPS file. The first two files must share the same time units and time steps, the DMPS file must also be with constant time interval.

Use only file name

Checking this option will ignore any directory name before the filename in the three input files and look for the file in directory "INPUT", located in the same directory where the "Current output path" is (it is shown in the notification area in the bottom of the program window). This option is not there to confuse you but to help portability. For example, if your input files for different days or cases are all named similarly (might not be the best overall solution but this happens), they would have to be located in different location. This way the files could be copied in to INPUT directory and the model would always look from that (relative) location. This can be especially relevant to in handling input data for batch cases.

Create batch

This tool is explained here: Create batch of simulations.

Time options

Runtime

Enter the simulation time either in hours or seconds.

Print every / Save every

• "Prints" are the terminal outputs, that helps you to check that everything is going the way you want with the simulation. The prints can be viewed in the tab  Run ARCA→Monitor. here we set the interval (in seconds) for the output.
• "Saves" are the file saving events, so this is analogous to "Print every".

X Samples

This option will calculate a nice interval for the Save every which will produce X saved points between the first and the last time. To get a number that coincides with the timestep, the exact amount of saves could still vary by a some amount.

Model timestep

This will define the fixed timestep, or, if the Precision based timestep is used, the minimum timestep. The model will increase the timestep if the precision targets can still be met, but it will not go under the minimum timestep, then the precision targets are omitted if necessary.

Precision based timestep

This lets the model vary the timesteps of the individual processes so that the predefined conditions are met. These conditions are measured by the change of some variable in one integration timestep. If this mode is used, the Model timestep can and should be set to quite low number, for example 0.001 seconds (this is done automatically when you turn on time step optimization).

• dDp, Particle diameter maximum and minimum relative change in percentages in one integration time step.
• dNp, Particle number concentration maximum and minimum relative change in percentages in one integration time step.
• dC, Condensable vapour maximum and minimum relative change in percentages in one integration time step.

The changes are saved in the file Changes.txt with the Interval defined in "Save every xx seconds" option. NOTE that these changes are just a snapshot, there might be larger changes between these samples. Also, remember that the values in Changes.txt are not a direct indication of the strength of some process, unless constant time step was used, in which case the values are comparable with each other throughout the run. Otherwise, if optimized timestep was used, at one point a 1% change could have happened in 0.5 seconds, whereas on another instance there was 2% change in 120 seconds. So be careful not to interpret the values in Change.txt as anything else than as indication if the chosen timestep was reasonable!

NOTE: The chemistry module is using Rosenbrock integrator and is able to handle larger time steps than aerosol module's Euler forward. Therefore when both chemistry and condensation is in use, the chemistry time step is governed by condensation. If, however, you use the chemistry module and NOT condensation, the changes in concentration (dC) are checked against chemistry, and these could be quite large. In order to get stable runs you need to use larger tolerances, setting the upper value around 10-50%, and use very small minimum time step. This will not mean slow runs, only that there is enough flexibility available for the time step.

What if the tolerances are not met?

In that case the time step will stay in the minimum time step, defined in Model time step. After 1000 rounds, if the changes are still above the tolerances, the model will save the output files and quit. This is to ensure that the model does not get caught in (nearly) infinite loop.

Maximum timestep and Nyquist theorem

The maximum timesteps for chemistry and condensation, coagulation and deposition is by default 150 seconds for all. If the model is used with such long timesteps, keep in mind the Nyquist theorem; to reproduce a signal you have to use at least twice as fast sampling frequency as the signal frequency. This is relevant in the point of view of the time dependent input variables. To manually change the maximum dt, insert in Custom option (Tab Run ARCA→ Custom model options) SPEED_DT_LIMIT <chem&cond>, <coag>, <depos>. The commas between are essential. For example here the limits have been set to 30 seconds:

Modules in use

Use chemistry: This is the KPP module, producing all the concentrations of the chemical reactions. You need to define your chemistry scheme first, otherwise you are resorting to the default one, which might or might not suit your purposes. See instructions how to define your chemistry schemeCreate chemistry scheme.

Form new particles: Anything that produces formation rates of new particles. Note that even if this option is off, you can still send in NUC_RATE_IN in the Time dependent input.

Use Aerosols: simulates aerosol dynamics

Aerosols include condensation: vapours from the chemistry module can condense and evaporate on the particles. Condensible vapours are defined in the Vapour file, defined in tab Aerosol→ Condensing vapours

Aerosols include coagulation: this turns on and off the simulation of coagulation.