Error message

Deprecated function: The each() function is deprecated. This message will be suppressed on further calls in _menu_load_objects() (line 579 of /home/cnrcwp/cnrcwp_drupal/includes/

Do RCMs improve representing North America Storm Tracks?

We are involved in the theme A3 of the CNRCWP project that seeks to understand current and future changes in weather storminess over Canada. In this note, I am presenting a brief assessment of the fifth generation Canadian Regional Climate Model (CRCM5) with respect to winter season extratropical cyclone (EC) characteristics over the current climate (1981-2005). This approach is also interested in evaluating the RCM added-value relative to its driving simulations which are in our case the ERAI reanalysis and two Global Climate Models (GCMs). The NCEP regional reanalysis (NARR) is used as reference simulation for bias calculations. A Lagrangian storm tracking algorithm is used to retrieve EC trajectories and characteristics. For more details, please read our recent paper Poan et al. (2017,

When CRCM5 is driven by ERAI, no significant skill deterioration arises and, more importantly, storm genesis (not shown) near areas with marked relief and their occurrence over continental lakes and onto the Atlantic are significantly improved with respect to ERAI. Conversely, in GCM-driven simulations, the added-value contributed by CRCM5 is less prominent and not consistent (from one realization to another), except over western NA areas with high topography. Over the Western Atlantic coastlines where the most frequent and intense ECs are located, significant intensity biases are found (figure 1). The inherited GCM sea surface temperature biases are thought to transmit unfavorable conditions to RCM performance over this area. Our ongoing work is focusing on analysing those East Coast ECs and particularly their link with precipitation.

Figure 1: 25-yr [1981-2005] NDJFM climatology of storm intensity bias in CVU (cyclonic vorticity units, 1 CVU = 10-5s-1) with respect to NARR and ERAI. (a) Driving data (ERAI, CanESM2 and MPI-ESM-LR and (b) CRCM5. In each case (a or b), the top row represents biases relative to NARR and the bottom represents biases relative to ERAI. Statistically significant differences (biases) in the sense of a Mann–Whitney–Wilcoxon test with p-value<0.01 are shaded by black dots.

Add new comment