Skip to main content
Earth, Planets and Space

Table 2 Results of validation with RAOBs

From: Comparison of different machine learning approaches for tropospheric profiling based on COSMIC-2 data

 

Artificial Neural Network

Random Forest

CDAAC wetPf2

Bending angle

Refractivity

Bending angle

Refractivity

Temperature [K]

1.89

1.81

2.16

2.01

1.29

Pressure [hPa]

1.93

1.83

2.13

1.89

1.35

Water vapor pressure [hPa]

0.47

0.46

0.54

0.52

0.41

  1. Vertically averaged RMSE between the temperature, pressure, and water vapor partial pressure obtained using different inputs and machine learning models, operational CDAAC wetPf2 and 477 co-located RAOBs with co-location criteria of 70 km and 2 h time window
Back to article page