2. Setting the mapper with output concentrations

Here, we propose to extract output concentrations from pre-computed output files. This is possible in the template by using the option only_outputs in the model paragraph.

 1rootdir: &rootdir /tmp/CIF/
 2outdir: &outdir  !join [*rootdir, /.tox/py38/tmp]
 3verbose: 2
 4workdir:  !join [*outdir, /4_onlyconcs]
 5logfile: log.txt
 6datei: 2000-01-01 00:00:00
 7datef: 2000-02-01 00:00:00
 8model:
 9  plugin:
10    name: template
11    version: std
12  require_domain: true
13  require_chemistry: true
14  only_outputs: true
15mode:
16  plugin:
17    name: forward
18    version: std
19domain:
20  nlat: 15
21  nlon: 10
22  plugin:
23    name: dummy
24    version: std
25  xmax: 50
26  xmin: -10
27  ymax: 60
28  ymin: 40
29obsoperator:
30  plugin:
31    name: standard
32    version: std
33chemistry:
34  schemeid: template_CO2
35  acspecies:
36    CO2:
37      restart_id: 1
38      mass: 12
39  emis_species:
40    CO2: null
41datavect:
42  plugin:
43    name: standard
44    version: std
45  components:
46    concs:
47      parameters:
48        CO2:
49          duration: 2H22min
50          frequency: 1H
51          nstations: 1
52          plugin:
53            name: random
54            type: measurements
55            version: param
56          seed: true
57          zmax: 100

Compared to the previous yaml, the following elements have been added:

  1. mode paragraph: this is the execution mode. It is needed for pyCIF to go further than the initialization step. Here we are computing a forward simulation to compare simulations with observations

  2. chemistry paragraph: this paragraph indicates information about species to be transported and/or emitted in the model

  3. datavect paragraph: this paragraph indicated all required information about pyCIF inputs. Further details is given here

In the example above, observations are generated at random locations. If one needs/wants to use a file of its own making, please refer to the dedicated documentation page about observation inputs (here) to properly prepare the file.

Once successfully running, the forward working files are saved in the folder $runsubdir/obsoperator/fwd_0000/. In particular, the output simulations and corresponding observations are saved here: $runsubdir/obsoperator/fwd_0000/obsvect.