Table 3 The computed means of each goodness-of-fit statistics metric, including bias, RMSE, ubRMSE, and R in the evaluation against CEOP, GHCNd, AKAH, and TP snow depth measurements for both open loop (OL) and multivariate data assimilation (MVDA) estimates between 2003 and 2019. n denotes the total number of stations.

From: Development of a multidecadal land reanalysis over High Mountain Asia

Evaluation sources

vs. CEOP snow depth (n = 3)

vs. GHCNd snow depth (n = 32)

vs. AKAH snow depth (n = 87)

vs. TP snow depth (n = 97)

Model estimates

OL

MVDA

OL

MVDA

OL

MVDA

OL

MVDA

Bias (m)

0.01

0.23

0.06

0.04

0.26

0.32

0.022

0.016

RMSE (m)

0.04

0.26

0.15

0.19

0.40

0.44

0.044

0.037

ubRMSE (m)

0.04

0.09

0.10

0.10

0.30

0.29

0.038

0.033

R (−)

0.61

0.30

0.55

0.55

0.49

0.56

0.23

0.23

Back to article page