Format

This plugin reads the SMET files as specified in the MeteoIO pre-processing library documentation (under "Available plugins and usage", then "smet").

Note: There is also a python library, pySMET available, to read SMET files.

Fluxes timeseries

These files are very regular SMET files with a large number of fields.

Layers data

The snow/soil layers file has the structure described below:

the SMET signature (to identify the file as SMET as well as the format version)
a Header section containing the metadata for the location
a Data section containing description of the layers (if any). Please note that the layers are given from the bottom to the top.

The following points are important to remember:

the station_id field is often used to generate output file names
you don't have to provide lat/lon, you can provide easiting/northing instead alongside an epsg code (see MeteoIO documentation)
the ProfileDate will be used as starting date for the simulation. Therefore, make sure you have meteorological data from this point on!
the number of soil and snow layers must be right!
timestamps follow the ISO format, temperatures are given in Kelvin, thicknesses in m, fractional volumes are between 0 and 1 (and the total sum must be exactly one), densities are in kg/m³ (see the definition of the fields in the table below)

initial snow profile fields description

Field	Description
timestamp	ISO formatted time
Layer_Thick	layer thickness [mm]
T	layer temperature [K]
Vol_Frac_I	fractional ice volume [0-1]
Vol_Frac_W	fractional water volume [0-1]
Vol_Frac_WP	fractional preferential water volume [0-1]
Vol_Frac_V	fractional voids volume [0-1]
Vol_Frac_S	fractional soil volume [0-1]
Rho_S	soil density [kg/m3]
Conduc_S	mineral phase soil thermal conductivity [w/(mK)]
HeatCapac_S	mineral phase soil thermal capacity [J/(kg*K)]

Field	Description
rg	grain radius [mm]
rb	bond radius [mm]
dd	dendricity [0-1]
sp	spericity [0-1]
mk	marker, see Metamorphism.cc
mass_hoar	mass of surface hoar []
ne	number of elements
CDot	stress change rate (initialize with 0.)
metamo	currently unused

Usually, simulations are started at a point in time when no snow is on the ground, therefore not requiring the definition of snow layers. An example is given below with one snow layer (and some comments to explain the different keys in the header):

SMET 1.1 ASCII
[HEADER]
station_id       = DAV2
station_name     = Davos:Baerentaelli
latitude         = 46.701
longitude        = 9.82
altitude         = 2560
nodata           = -999                                ;code indicating a missing value
tz               = 1                           ;time zone in hours
source           = WSL-SLF                     ;optional key
ProfileDate      = 2009-10-01T00:00            ;when was the profile made, see explanations above
HS_Last          = 0.0000                      ;last measured snow height
SlopeAngle       = 38.0
SlopeAzi         = 0.0
nSoilLayerData   = 0                           ;number of soil layers
nSnowLayerData   = 1                           ;number of snow layers
SoilAlbedo       = 0.20                                ;albedo of the exposed soil
BareSoil_z0      = 0.020                       ;roughtness length of the exposed soil
CanopyHeight     = 0.00                                ;height (in m) of the canopy
CanopyLeafAreaIndex     = 0.00
CanopyDirectThroughfall = 1.00
WindScalingFactor       = 1.00                 ;some stations consistently measure a wind that is too low
ErosionLevel     = 0
TimeCountDeltaHS = 0.000000
fields           = timestamp Layer_Thick  T  Vol_Frac_I  Vol_Frac_W  Vol_Frac_V  Vol_Frac_S Rho_S Conduc_S HeatCapac_S  rg  rb  dd  sp  mk mass_hoar ne CDot metamo
[DATA]
2009-09-19T02:30 0.003399 273.15 0.579671 0.068490 0.351839 0.000000 0.0 0.0 0.0 1.432384 1.028390 0.000000 1.000000 22 0.000000 1 0.000000 0.000000

Hazard data

The hazards file contain the temporal history of various parameters that are relevant for avalanche warning (such as three hours new snow fall, etc). If such file is not provided, the internal data structures for such data will be initialized to zero (which is what you usually want when starting a simulation before the start of the snow season). The hazards file has the following structure:

SMET 1.1 ASCII
[HEADER]
station_id       = DAV2
station_name     = Davos:Baerentaelli
latitude         = 46.701
longitude        = 9.82
altitude         = 2560
nodata           = -999
tz               = 1
ProfileDate      = 2012-06-11T17:30
fields           = timestamp SurfaceHoarIndex DriftIndex ThreeHourNewSnow TwentyFourHourNewSnow
[DATA]
2010-06-08T18:00       -999       -999   0.000000   0.000000
2010-06-08T18:30       -999       -999   0.000000   0.000000
2010-06-08T19:00       -999       -999   0.000000   0.000000
2010-06-08T19:30       -999       -999   0.000000   0.000000
2010-06-08T20:00       -999       -999   0.000000   0.000000
2010-06-08T20:30       -999       -999   0.000000   0.000000
2010-06-08T21:00       -999       -999   0.000000   0.000000
...
2010-06-11T17:30       -999       -999   0.000000   0.000000

As can be seen in this example, the various indices as well as the snow statistics are given every half an hour in reverse chronological order until the profile date.