amdados.conf

##### Parameters for Amdados application.
##### This is the default settings used in all the tests.
##### Only the number of sub-domains and integration period are changed
##### during the testing.

output_dir  output      # the output folder, which should be created by user

diffusion_coef 1.0      # diffusion coefficient

### Domain geometry.
num_subdomains_x 4     # number of sub-domains in x-dimension
num_subdomains_y 2     # number of sub-domains in y-dimension

                        # Sizes that are hard-coded as C++ template parameters:
subdomain_x 16          # number of sub-domain nodal points in x-dimension
subdomain_y 16          # number of sub-domain nodal points in y-dimension

domain_size_x 1000      # global domain size in x-dimension [meters]
domain_size_y 1000      # global domain size in y-dimension [meters]

### Location and intensity of a contaminant spot at time t=0.
spot_x 30               # substance spot abscissa in the global domain [meters]
spot_y 150              # substance spot ordinate in the global domain [meters]
spot_density 10000      # substance spot concentration at initial time [units?]

### Sensors.
sensor_fraction  0.0025   # fraction of points occupied by sensors; \
                          # the fraction is only approximately safisfied.

# XXX Not used for now.
#sensor_per_subdomain 0    # 0 - sensors are placed pseudo-randomly across the
#                          # domain and a subdomain may have no single one.
#                          # 1 - sensors are places pseudo-randomly and
#                          # at least one sensor presents at each subdomain.

### Integration of advection-diffusion model.
integration_period 25   # integration period 0...T [seconds]
integration_nsteps 50    # min. number of integration time steps
num_sub_iter        3      # fixed number of sub-iterations on each step

### Parameters of flow in the domain.
flow_model_max_vx 1.0   # module of max. flow velocity in x-dimension [m/s]
flow_model_max_vy 1.0   # module of max. flow velocity in y-dimension [m/s]

### Kalman filter.
                              # Model covariance matrix P:
model_ini_var           1.0   # initial variance of diagonal elements of P
model_ini_covar_radius  1.0   # initial radius of correlation between \
                              #                    domain points [meters]

                        # Noise covariance matrices:
model_noise_Q    1.0    # model noise variance (a basic, reference value)
model_noise_R    1.0    # measurement noise variance (a basic, reference value)

### Storing the result/visualization/etc.
write_num_fields 100    # record this number of full fields during simulation

### Testing and debugging

kalman_time_gap 1      # invoke Kalman filter every 'gap' time step,
                       # N O T E, normally this parameter must be equal 1,
                       # and only in the special testing mode we set it
                       # greater than 1.