-
Notifications
You must be signed in to change notification settings - Fork 5
Model parameters and xml parameter files
Settings in the ED2 setting file correspond to variables defined in the “ed_data” object. These variables have been lumped into a number of groupings related to their role in the model and the common block memory they are assigned to:
| Misc. Settings | <misc></misc> |
| <ed_misc></ed_misc> | |
| Plant Functional Types | <pft></pft> |
| PFT constants | <pftconst></pftconst> |
| Hydrology | <hydro></hydro> |
| Lapse Rates | <lapse></lapse> |
| Solar Radiation/Light Interception | <light></light> |
| Soils | <soil></soil> |
| Decomposition | <decomposition></decomposition> |
| Patch/Cohort Fusion/Fission | <fusefiss></fusefiss> |
| Disturbance | <disturbance></disturbance> |
| Phenology | <phenology></phenology> |
| Physiology | <physiology></physiology> |
| Additional XML files | <extern></extern> |
Configuration files are encoded in XML format using the above tags. The parser doesn't care about the order or number of tags in the config file. In fact, one must specify one <pft></pft> section per plant functional type. Currently there are a large number of defaults built into the model. The config file is read AFTER all the defaults have been set, so any values not set in the config stay at their default. Within a section, as few or as many values can be set as the user desires. An example of a config file might be:
<?xml version=”1.0”>
<config> <ed_misc> <restart_target_year>1993</restart_target_year> </ed_misc> <pft> <num>5</num> <name>SouthernPine</name> <SLA>10</sla> </pft> </config>
Since XML can be easily read/written by a large number of programs/languages, the use of XML config files makes it easy to generate “batch jobs” or “ensembles” of runs where parameters can be sampled from, as is done in http://pecanproject.org https://github.com/PecanProject/pecan/tree/master/models/ed
Because they were large tables, each section has been moved to it's own page:
In addition to parameters that apply to the model as a whole, there are parameters that vary in space. Foremost among these are the parameters for soil physics. The soil physics for a polygon are set using two databases. First, the soils are set using a global dataset from the UN Food and Agriculture Organization (FAO) at a 1x1 degree resolution that has been converted to a HDF5 file. Second, soil physics is updated, where available, from the ~1x1 km USGS STATSGO soils database. The directory and prefix for these soil texture files are set by ISOILFN. The STATSGO soils are stored in ASCII files in 5x5 degree tiles that are generated using the program “rextract” which reads the data from the binary file “layertext.bsq”. This database has not yet been ported to v2.1 using HDF5. Data in each file is stored in a 2 character field (no delimiter) and takes on the following values:
| 1 | sand | 6 | Sandy clay loam | 11 | Clay | 16 | Clayey sand |
| 2 | Loamy sand | 7 | Silty clay loam | 12 | Peat | 17 | Clayey silt |
| 3 | Sandy loam | 8 | Clay loam | 13 | Bedrock | ||
| 4 | Silt loam | 9 | Sandy clay | 14 | Silt | ||
| 5 | Loam | 10 | Silty clay | 15 | Heavy clay |
Soil physical parameters are then assigned based on a table. The format of the file is with longitude as the outer loop, latitude the middle loop, and soil depth the inner loop. The code assumes 11 soil layers with pre-defined depths (centers = 0.025,0.075,0.15,0.25,0.35,0.5,0.7,0.9,1.25,1.75,2.25m) and that the lat/lon grid is 600x600 cells (i.e. the soils are actually stored on a 30 x 30 second grid). Currently the modal soil type (i.e. The type with the highest frequency) is assigned to the whole polygon, however the soil types for individual types can be specified in the site file, as described in Initial Conditions.