Evaluation of the CABLEv2.3.4 Land Surface Model Coupled to NU‐WRFv3.9.1.1 in Simulating Temperature and Precipitation Means and Extremes Over CORDEX AustralAsia Within a WRF Physics Ensemble

Abstract

The Community Atmosphere Biosphere Land Exchange (CABLE) model is a third-generation land surface model (LSM). CABLE is commonly used as a stand-alone LSM, coupled to the Australian Community Climate and Earth Systems Simulator global climate model and coupled to the Weather Research and Forecasting (WRF) model for regional applications. Here, we evaluate an updated version of CABLE within a WRF physics ensemble over the COordinated Regional Downscaling EXperiment (CORDEX) AustralAsia domain. The ensemble consists of different cumulus, radiation and planetary boundary layer (PBL) schemes. Simulations are carried out within the NASA Unified WRF modeling framework, NU-WRF. Our analysis did not identify one configuration that consistently performed the best for all diagnostics and regions. Of the cumulus parameterizations the Grell-Freitas cumulus scheme consistently overpredicted precipitation, while the new Tiedtke scheme was the best in simulating the timing of precipitation events. For the radiation schemes, the RRTMG radiation scheme had a general warm bias. For the PBL schemes, the YSU scheme had a warm bias, and the MYJ PBL scheme a cool bias. Results are strongly dependent on the region of interest, with the northern tropics and southwest Western Australia being more sensitive to the choice of physics options compared to southeastern Australia which showed less overall variation and overall better performance across the ensemble. Comparisons with simulations using the Unified Noah LSM showed that CABLE in NU-WRF has a more realistic simulation of evapotranspiration when compared to GLEAM estimates.

Atmospheric and Land Surface Contributions to Heatwaves: An Australian Perspective

“…For each case study, an offline spinup of 30 years of the hydrological state variables within CABLE was performed using the 3‐hourly Princeton meteorological forcing (Sheffield et al, 2006). Further details of the CABLE land surface model configuration used here is available in Hirsch, Kala, et al (2019).…”

“…Through the NASA Land Information System (LIS; Kumar et al, 2006;Peters-Lidard et al, 2007) the Australian Community Atmosphere Biosphere Land Exchange (CABLE) land surface model is coupled to NU-WRF. This coupled model environment is referred to as WRF-LIS-CABLE (Hirsch, Kala, et al, 2019). WRF-LIS-CABLE has been used previously to understand land-atmosphere interactions over Australia focusing both on extremes (Hirsch et al, 2014) and sensitivity to land cover change (Hirsch et al, 2015).…”

Mapping groundwater abstractions from irrigated agriculture: big data, inverse modeling, and a satellite–model fusion approach

“…In our study, a coupled leaf-canopy model (PROSAIL) produced forward runs over a wide range of realizations, and these were used as a training dataset to develop estimates of LAI using a random forest (RF) approach. The inversion of the forward runs needed to derive LAI was based on the REGularized canopy reflectance model (REGFLEC; Houborg et al, 2015), which has been shown to be suitable for largely automated applications (Houborg and McCabe, 2016). The configuration of REGFLEC in this study followed Houborg and McCabe (2018a) but using only the model inversion results (e.g., PROSAIL) due to a lack of in situ LAI data in the larger region of Al Jawf.…”

Could crop albedo modification reduce regional warming over Australia?

“…All simulations were carried out using the Noah land surface model (Chen and Dudhia, 2001), which is the most commonly used and extensively evaluated land surface model in WRF. Given that WRF is sensitive to physical parameterization options over both SEA and SWWA (Evans et al, 2012;Hirsch et al, 2019b;Kala et al, 2015a), we ran three ensembles to sample model structural variability (Table 1) based on these studies. The first ensemble used the Yonsei University planetary boundary layer (PBL) scheme with Monin-Obukhov similarity for the surface layer (Hong et al, 2006), the Rapid Radiative Transfer Model (RRTM) scheme for longwave radiation (Mlawer et al, 1997), the Dudhia scheme for shortwave radiation (Dudhia, 1989), and the Kain-Fritsch scheme for convection (Kain, 2004).…”

“…The second ensemble differs from the first by using the Mellor-Yamada-Janjic PBL scheme with ETA scheme for the surface layer (Janjić, 1994), and the third ensemble differs from the first ensemble by using the Betts-Miller-Janjic scheme for convection (Betts and Miller, 1986). We do not carry out additional model evaluation against observations in this paper, as all these WRF configurations, which all use the Noah land surface model, have already been extensively evaluated over both SEA and SWWA by numerous studies against maximum and minimum temperature and precipitation observations, and shown to reproduce the observed climatology reasonably well (Andrys et al, 2016(Andrys et al, , 2015Di Virgilio et al, 2019;Evans et al, 2012;Evans and McCabe, 2010;Firth et al, 2017;Hirsch et al, 2019aHirsch et al, , 2019cKala et al, 2015a).…”