GRIB-API definition

NamepercolationAbbreviationpercUnitkg m-2 paramId:260430


UERRA details

DefinitionThe percolation is the downward movement of water under hydrostatic pressure in the saturated zone.
This water might still end up in rivers and lakes as discharge but it is a slower process than water runoff or drainage.
Such defined percolation is an input for hydrological models together with e.g. water runoff.
Validityaccumulated  (from the beginning of the forecast)
Comment



WMO GRIB2 definition

Parameter
Discipline1hydrological products
Parameter category0hydrology basic products
Parameter number16percolation rate**
Type of statistical processing1accumulated

**newly approved  by WMO (release in May 2016)


Level
Type of first fixed surface177Deep soil (of indefinite depth)
Scale factor of first fixed surfaceall bits to 1missing
Scaled value of first fixed surfaceall bits to 1missing
Type of second fixed surfaceall bits to 1missing
Scale factor of second fixed surfaceall bits to 1missing
Scaled value of second fixed surfaceall bits to 1missing

9 Comments

  1. @Sebastien Villaume

    Sent: Wednesday, 24 June, 2015 1:31:21 PM

    Subject: Re: urgent quick help - some MF surface parameters

    > drainage (separate from runoff which is another used parameter..)

    well, WMO has only one parameter for both drainage and run off, Discipline 2 Product Category 0 - Parameter 33. My guess is that the same parameter should be use but on 2 different types of Level, like above or below surface or something. I am not an expert but I thought that runoff occurs above the surface and drainage is water taken from under surface? Or?

  2. There is a special level  106 "Depth below land surface" in GRIB2 but how to specify the bottom level for drainage?

    Or maybe there is some special type of subsurface layer defined which I missed but could be used?

  3. There are in total 6 different runoff definitions by WMO (some hydrological) which is quite excessive. The definition of runoff is water that leaves the the soil/ground and becomes surface water.

    If we stick to the land surface (Disc 2, cat 0), Number 5 is Water runoff, so that should be used for normal runoff. Parameter 33 could be used for the drainage I assume, since you have used Param 34 for Surface water runoff. We can have a chat about this tomorrow morning.

    1. I agree very much with Fredrik's proposal above. There is a merit of having the two different parameter id:s as above.
    2. The validity "instantaneous" is not correct I think, but it should be accumulated (since the start of the forecast usually).
    3. Level definition for drainage I don't think is needed, since there is never going to be more than one drainage (OK?) so I don't think one has to hunt a lot for finding "the bottom" or similar, but of course if there is something like that, then use it, or just stick it in the "Comment".
  4. Eric Bazile (Sent: Thursday, 15 October, 2015 2:36:35 PM)

    Runoff is the water going out the surface reservoir (after precipitation) : lateral path (water is lost for the model). Drainage is the water at the bottom of the deepest layer : vertical. Both are used as input for hydrological model. For runoff and drainage: accumulated Kg/m2 for one hour. I think, in general, for fields available hourly, the accumulated period will be 1 hour.

  5. Ok, so drainage in this sense is what in hydrology is called percolation. This water is still interacting with the water that ends up in rivers and lakes as discharge, but it is a slower process.

  6. If percolation is really what MF produces and should go into UERRA archive we should propose that as a new parameter to WMO as the currently existing WMO "Water runoff and drainage"  does not seem to be the same thing.

    1. I think this is a bit overboard to use a different parameter for one model e.g. since it seems to me that it is mainly a question of drainage from the meteorological (soil) model. Then the hydrological modelling would handle it differently
    2. I still wonder if and why we can't have the two different parameter id:s as proposed by Fredrik above. It looks odd to me to have the two names with an "and" since it seems to be the only case here and they are two physically different things (at different levels and with lateral transports being fast or resp. slow.