high resolution snow analysis for cosmo n.
Skip this Video
Loading SlideShow in 5 Seconds..
High Resolution Snow Analysis for COSMO PowerPoint Presentation
Download Presentation
High Resolution Snow Analysis for COSMO

Loading in 2 Seconds...

play fullscreen
1 / 19

High Resolution Snow Analysis for COSMO - PowerPoint PPT Presentation

  • Uploaded on

High Resolution Snow Analysis for COSMO. jean-marie.bettems@meteoswiss.ch COSMO General Meeting 18-21 September 2007. 20040310. Satellite data. Near real time, high resolution, composite, partial snow cover Based on: Meteosat SEVIRI. 20070325. Snow depth analysis .

I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
Download Presentation

PowerPoint Slideshow about 'High Resolution Snow Analysis for COSMO' - duke

An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript
high resolution snow analysis for cosmo

High Resolution Snow Analysis for COSMO


COSMO General Meeting18-21 September 2007

satellite data


Satellite data

Near real time, high resolution, composite, partial snow cover

Based on: Meteosat SEVIRI

snow depth analysis


Snow depth analysis

Near real time, high resolution, snow depth anaysis

Based on: in-situ observations, Meteosat mask, COSMO model

  • Fractional snow cover is derived from satellite automatically, in near-real time, at 2 km resolution.
  • A snow depth map is produced daily over western and central Europe on a 2.2 km grid.
  • both are unique
  • snow depth map is more realistic than current products; Meteosat information generates more realistic small scale structures by adding or removing snow patches
  • improved COSMO near surface weather in winter (e.g. 2m T)
summary deliverables
Summary - Deliverables
  • Snow depth analysis for COSMO-7 in production
  • Snow depth analysis for COSMO-2 in pre-production
  • Scientific (EUMETSAT final report) and technical documentation available
satellite data1
Satellite data

General problems:

  • obscurance of the surface by clouds
  • confusion of ice clouds and snow (similar spectral signatures)


  • high temporal frequency  MSG SEVIRI

EUMETSAT Fellowship:

  • detect dynamic behaviour of clouds for improving the discrimination between clouds and snow (with respect to spectral classification alone)
  • detect all cloud-free instances to reduce obscurance of surface by clouds
  • map snow cover automatically and in near-real time


satellite data2



Satellite data

SEVIRI characteristics

  • Coarse to medium spatial resolution: 5-6 km and 1.5-2 km (HRV)
  • High temporal resolution: each 15 minutes, only day-time images used
  • Adequate spectral resolution: 12 spectral channels, 10 used:

1 VIS 0.635 m

2 VIS 0.81 m

3 NIR 1.64 mm

4 IR3.92m

5 IR 6.2 mm

6 IR 7.3mm

7 IR 8.7 mm

8 IR 9.7 mm

9 IR10.8 mm

10 IR12.0 mm

11 IR13.4 mm

12 HRV 0.7 mm


Satellite data

Classification scheme:

multi-channel colour composite

(red: snow or ice clouds)

temporal standard deviation, channel 3

(dark: low; bright: high)

classification result

(white : snow; dark gray : clouds)


Satellite data

  • Snow cover products:
  • instantaneous snow map
  • daily composite snow map: all cloud free instances from 1 day combined
  • running composite snow map: continuously updated with the latest cloud-free information (each pixel displays the latest instance that the pixel was cloud-free)
  • quality flag taking into account snow depth at time of occultation
  •  q=f(time,sza,n)
  • Properties:
  • fully automatic processing in near-real time (new image processed 2.5 hours after acquisition, each 15 minutes)
  • fractional snow cover
  • normal SEVIRI resolution (5-6 km) and high SEVIRI resolution (1.5-2 km)
satellite data3
Satellite data


20040310 13:12 UTC

20040308-20040310, composite snow fraction

20040310 13:12 UTC, snow fraction

20040308-20040310, composite quality

satellite data4
Satellite data


high resolution

normal resolution




Satellite data


  • winter 2005/2006:
    • normal resolution: 94% correlation with in situ observations
    • consistent quality over the whole period
  • March + April 2007:
    • normal resolution: 95% correlation (only Alps: 83%)
    • high resolution: 96% correlation (only Alps: 87%)
snow analysis method
Snow analysis: method

DWD software package for computing snow depth maps, adapted and optimised at MeteoSwiss for use with MSG SEVIRI.

  • Cressman analysis
    • Interpolation between observations of snow depth
  • depending on observation density:
    • use interpolated snow depth only,
    • or add interpolated precipitation
    • or add model snow depth
  • compare with satellite data
    • always use latest version of running composite SEVIRI snow map
    • resample SEVIRI snow map to model space
    • only use satellite information that has high quality
    • remove/add snow from Cressman analysis to match SEVIRI information

Current data set

  • mainly synop
  • sparse

Potential additional data set

  • considerably more data, but …
  • … several data providers
  • … several data formats

In-situ observations

snow analysis alternative interpolation method
Snow analysis: alternative interpolation method

Snow depth generally increases with surface altitude  use local gradients for


for each model grid point (x,y,z):

  • find np observation sites with smallest distance ,

only use sites within horizontal distance Rmax and within vertical distance Hmax

  • make linear regression for these sites: snow depth = a + b z

(weight the contribution of each site with the inverse of d)

  • use this regression line to compute the snow depth at (x,y,z)

(only when enough of the np sites display snow, e.g. half of them)

snow analysis alternative interpolation method1
Snow analysis: alternative interpolation method

Cressman analysis

altitudinal interpolation

(note: different geographic projection, no influence of satellite data)

  • Merge latest DWD snow analysis modifications with new software
  • Access to additional in-situ observations
    • Currently DWD data can not be decoded
  • Interpolation with altitudinal gradient
    • more realistic than Cressman interpolation over steep topography, but enough observations with snow must be present
    • use gradient-interpolation to identify bad observations
    • merge gradient-interpolation with cressman analysis, e.g. with weighted mean
  • Use partial snow cover in COSMO/TERRA
  • Use EUMETSAT SAF snow albedo in COSMO
  • Introduce a more sophisticated snow model