MET Nordic analysis operational OI settings

[tnipen]

At MET Norway, we produce the MET Nordic Analysis product, which is produced with hourly resolution at 1 km. More information about the product is available here: https://github.com/metno/NWPdocs/wiki/MET-Nordic-dataset

In this product, we use OI for 3 different products:

Air temperature analysis

• Background model: MEPS ensemble downscaled with dynamic elevation gradient to 1.0 km
• Observations: SYNOP, Netatmo, ++
• OI function: optimal_interpolation_ensi
• Structure function: Barnes
• Horizontal length scale: 10 km
• Vertical length scale: 100 m
• Land area fraction length scale: 0.85
• Max points: 20
• Sigma: 0.5
• allow_extrapolation: False
• observation operator: Elevation gradient (6.5°C/km)

Precipitation analysis
OI is run sequentially 5 times, with a horizontal length scale of 10km, 8km, 6km, 4km, 2km

• Background model: MEPS ensemble at 2.5 km
• Observations: SYNOP, Netatmo, ++
• OI function: optimal_interpolation_ensi
• Structure function: Barnes
• Horizontal length scale: 10, 8, 6, 4, 2 km
• Vertical length scale: None
• Land area fraction length scale: None
• Max points: 20
• Transformation: Box-cox (threshold = 0.5)
• Sigma: 1
• allow_extrapolation: False
• observation operator: Nearest neighbour

Cloud area fraction analysis

• Background model: MEPS control run interpolated bilinearly to 1.0 km
• Observations: Cloud cover estimates from webcamera images
• OI function: optimal_interpolation
• Structure function: Barnes
• Horizontal length scale: 20 km
• Vertical length scale: None
• Land area fraction length scale: None
• Max points: 20
• Error ratios: Varies based on estimated accuracy of observation
• allow_extrapolation: False
• Observation operator: Bilinear interpolation