NCI records for Errata is given in Errata and Known Issues for CORDEX-CMIP6 and Regional Downscaling - CMIP Community - Opus - NCI Confluence.
The Bureau of Meteorology seeks user feedback on the quality and their usage of the data, to help identify areas for improvements.
Please email help@nci.org.au
| Affected Model | Affected Domain | Affected Variable | Affected Frequency | Severity | Issues | Action | Status |
|---|---|---|---|---|---|---|---|
| BARRA-R2, BARRA-RE2 | AUS-11, AUS-22 | CAPE, CIN | All | Moderate | Biases in CAPE and CIN due to inaccuracies in parcel ascent computation in Unified Model UM v11.9. | Alternative data from Bureau's ConvParams will be made available but over a truncated domain limited to Australia | Resolved See FAQ |
| BARRA-C2 | AUST-04 | CAPE, CIN | All | Moderate | Biases in CAPE and CIN due to inaccuracies in parcel ascent computation in Unified Model UM v13.0. | Alternative data from Bureau's ConvParams dataset will replace this data | Current |
| BARRA-R2 | AUS-11 | tsl | day | Minor | Soil temperature over Mount Cook, New Zealand is cooler during 2013 than other years. The issue appeared isolated over 3 grid cells over Mt Cook. | None | Current |
| BARRA-R2, BARRA-RE2 | AUS-11, AUS-22 | mrsol(lev=2m), mrso, mrsol(lev=2m), mrso, mrsol(lev=2m), mrso | All | Moderate | Soil moisture content in the lowest layer or in the total column shows abrupt discontinuities on 1 September, which correspond to where the data from one production stream is combined with the next. | None | Current |
| BARRA-R2, BARRA-RE2 | AUS-11, AUS-22 | husN, taN, vaN, uaN, waN, zgN for N=1000, 950, 925, 850, 700 (units: hPa) | All | Minor | Pressure level data: Where pressure level is less than surface pressure, data has been masked with fill_value. The values of zg can fall to negative at few isolated locations, where the pressure is less than surface pressure. | None | Current |
| BARRA-R2, BARRA-RE2 | AUS-11, AUS-22 | huss, hurs, tasmean, tasmax, tasmin | 1hr | hu*: Moderate ta*: Minor | Spurious high values of humidity, due to the UM diagnostic calculation of 2m specific humidity. High values should be screened before use or 10m humidity should be substituted. A reduction in diagnosed screen temperature accompanies this spurious increase in diagnosed screen-level moisture. | 10 m humidity based on unaffected model-level fields over land will also be provided | Fixed in v20240516 for huss/hurs No further action required |
| BARRA-RE2 | AUS-22 | hus10 | 3hr | Minor | Drier than nominal on time step 2014-07-15 21:00 over SE regions of the domain for some ensemble members. | None | Current |
| BARRA-R2, BARRA_RE2 | AUS-11, AUS-22 | ta150m, ua150m, va150m, ta200m, ua200m, va200m | All | Minor | Masked in regions with high orography | None | Current |
| BARRA-R2, BARRA_RE2 | AUS-11, AUS-22 | mrfso, mrfsol, mrsol, mrsol | All | Moderate | Wet bias in the soil moisture over grid cells with urban surface type - all Australian cities, from January 2021. | Planning for re-running BARRA-R2/RE2 cycles from January 2021 | Fixed in v20240516 |
| BARRA-R2, BARRA_RE2 | AUS-11, AUS-22 | evspsbl | All | Major | Data is yet to be released. | Will be released in 2024 | Released in v20240516 |
| BARRA-R2, BARRA_RE2 | AUS-11, AUS-22 | prhmax | day, mon | Major | Data is yet to be released. | Will be released in 2024 | Released in v20240516 |
| BARRA-R2, BARRA-RE2, BARRA-C2 | AUS-11, AUS-22, AUS-04 | All | All | Moderate | The first two months of 1979, January and February 1979, is the spin-up months for the 1979 reanalysis so we advise caution when using the data from this two months. | None | Current |
| BARRA-RE2 | AUS-22 | wa200, wa300, wa400, wa500, wa600, wa700, wa850, wa925, wa1000 | All | Minor | Modelling artifact in ctl_0 ensemble member resulting in anomalously low value pixel off the coast of New Zealand at 1995-02-07 03Z and 04Z. Most obvious in wa1000. | May consider re-running the cycle | Current |
| BARRA-RE2 | AUS-22 | ta10, ta20, ta30, wa10, wa20, wa30 | All | Minor | Discontinuities in field mean apparent for low-pressure ta, wa from 2000 to 2008. Other pressure levels and variables (e.g. ua10, va10) are also affected to a lesser degree. Discontinuity is caused by ERA5.1 - a rerun of ERA5 from 2000 to 2006, which is expected to be "problematic for global-mean upper stratospheric temperature". | None | Current |
| BARRA-R2, BARRA-RE2, BARRA-C2 | AUS-11, AUS-22, AUST-04 | evspsblpot | 1hr, day | Minor | Very isolated grid cells showing larger negative or positive values at a single time step. | None | Current |
| BARRA-R2, BARRA-RE2, BARRA-C2 | AUS-11, AUS-22, AUST-04 | evspsblpot | 1hr, day, mon | Moderate | The potential evapotranspiration (PET) is the ET that would occur if the soil and vegetation are saturated. The PET scales with the friction velocity for each tile type (e.g., grass, shrub, broadleaf tree, bare soil), which depends on the canopy heights etc. PET is computed by the model for each tile type and the average across the tiles are computed and published. However, the energy constraint is not imposed on the calculation of this diagnostic in the model, which is different from using the Penman-Monteith equation. In other words, the PET diagnostic is simply driven by a surface to atmosphere humidity difference. Accordingly, it is expected that the PET diagnostic will be higher than the values in other PET products such as SILO PET (https://data.qld.gov.au/dataset/silo-climate-database). Different definitions of PET can be computed from BARRA's other gridded fields. For instance, Penman-Monteith equation requires 2m temperature (tas), relative humidity (hurs), wind speed (sfcWind), surface pressure (ps), net SW radiation at surface (rsds, rsus or rss), cloud cover (clt), | None | Current |
| BARRA-R2, BARRA-RE2 | AUS-11, AUS-22 | tauu, tauv | All | Moderate | The lat/lon grids for these variables remain in their modelling staggered grids, and thus inconsistent with other variables in the dataset | Reprocess the data to the common lat/lon grids | Fixed in v20231001 on |
| BARRA-R2, BARRA-RE2 | AUS-11, AUS-22 | Many | All | Minor | The lat/lon values of their grids are not consistently rounded up to two decimal points. | Rounding of the lat/lon values in process. | Fixed in v20240516 |
| BARRA-R2, BARRA-RE2 | AUS-11, AUS-22 | husX, taX, uaX, vaX, waX, zgX where X is 1000, 950, 925, 850, 700, 600, 500, 400, 300, 200, and omega500 | day, mon | Minor | Daily and monthly data for these variables were calculated using means of instantaneous fields every 3 hours instead of means of instantaneous fields every 1 hour. This can be improved by recalculating these fields using the 1hr data. Users are advised to use the 1hr data to derive daily and monthly data for more accurate estimates. | Recalculate these fields using the 1hr data. | Fixed in v20240613 |
| BARRA-R2, BARRA-RE2, BARRA-C2 | AUS-11, AUS-22, AUST-04 | All | All | Minor | Discontinuities in the statistics of some derived convective parameters were discovered during the period from 2015-09 to 2015-12. | Rerun BARRA2 for the affected period. | Fixed in v20240809 on |
| BARRA-RE2 | AUS-22 | 1hr/some, 3hr/some, day/*, mon/* | All | Minor | The affected data files contain "time_bnds" variable, which has dimensions "realization,time,bnds". The order of the dimensions, where "realization" comes before "time", prevent the files to work with CDO and NCO. | Remove "realization" from the dimensions of "time_bnds" | Fixed |
| BARRA-R2/RE2 | AUS-22 | Some | All | Minor | Some data files that should have "time_bnds" are missing them. | Add "time_bnds" as needed | Current |
| BARRA-R2, BARRA-RE2, BARRA-C2 | AUS-11, AUS-22, AUST-04 | wsgsmax | All | Moderate | An error is found in the Unified model version 11.9 to 13 in its calculation of the wind gust diagnostics. An incorrect value is used to down weight the gust contribution where there was a large orographic roughness length from the form drag roughness scheme, leading to positive biases. The biases are larger over land than over sea, and the preliminary experiments over UK show an averaged mean error over land is 1-1.5 m s-1, solely due to this error. On average, the mean error is around 10% over land, and 5% over sea. | None | Current |
| BARRA-C2 | AUST-04 | ztp | 1hr | Minor | The troposphere height is determined by checking the lapse rate of surrounding model height intervals. The height cannot always be determined through this approach for BARRA-C2 with 40 km model lid, and as such the height values are masked. | None | Current |
| BARRA-C2 | AUST-04 | MUCAPE | 1hr | Minor | Occasional grid cells have anomalously large values due to modelling artefacts (e.g. near the NT coast on 2008-03-07T13Z, where one grid cell has a value of 29e3 J kg-1 instead of the typical fieldmax of 12e3 J kg-1). | None | Current |
| BARRA-C2 | AUST-04 | taXm | All | Minor | Occasional grid cells have anomalously large values due to modelling artefacts (e.g. off the coast of Queensland in ta150m at 2014-01-26T05Z, where one grid cell has a value of 340K instead of the typical fieldmax around 323K). | None | Current |
| BARRA-C2 | AUST-04 | rsus | All | Minor | There is a grid cell with an anomalously low value, due to modelling artefact, off the coast of NSW in at 1983-07-08T2030Z. This grid cell has a value of -18.7 W m-2 instead of the typical fieldmin around -1.7 W m-2. | None | Current |
| BARRA-R2, BARRA-RE2, BARRA-C2 | AUS-11, AUS-22, AUST-04 | hus10/20/30 | All | Minor | Temporal discontinuities in the field mean of high altitude specific humidity are apparent in all datasets in early 1984, likely due to similar discontinuities observed in ERA5, see http://doi.org/10.1002/qj.3803 | None | Current |
| BARRA-C2 | AUST-04 | visibility | 1hr | Minor | Field minima are anomalously high (around 200m instead of the typical 50m minimum) for two cycles, 1979-11-13 00:00 & 1997-10-11 00:00. | None | Current |
| BARRA-R2, BARRA-RE2, BARRA-C2 | AUS-11, AUS-22, AUST-04 | All | All | Minor | BARRA-R2 incorrectly assimilated the 3m wind observations from Western Australia Department of Primary Industries and Regional Development (DPIRD) automatic weather stations as 10 m (standard meteorological station height) observations. This Concentrated over the southwest region of Australia around Perth, these stations report 3m anemeter heights. DA experiments using the 1.5 km NWP suite (ACCESS-PH) found that the effect of assimilating these wind obs appears to taper off after 1-2 hours forecasts, so thus they would have negligible impact on the BARRA2 products which are based on longer forecasts. The rejection of these 3m obs was implemented after 20240131T00Z cycle. | Action | Current |
| BARRA-C2 | AUST-04 | hfls, hfss, psl, sfcWind, ua*, va*, wa*, wap*, hus*, wsgs | All | Minor | BARRA2 follows the MetOffice's NWP data accuracy conventions, which results in occasional fields with observable quantisation of data values, particularly in some high altitude variables. E.g. in BARRA-C2 wa10 has a data resolution of 0.125 m/s and typically has a field mean of 0 with a data range <1 m/s. This effect is exacerbated with aggregated fields such as daily and monthly means. | None | Current |
| BARRA-C2 | AUST-04 | All | All | Minor | Occasionally anomalous values are present in isolated grid cells. The effects are observable across multiple variables and occasionally in following timesteps. The anomalies are more easily observed in the BARRA-C2 Convective Parameters sub-dataset but can still be detected in the primary BARRA-C2 variables. | None | Current |
| BARRA-R2, BARRA-RE2, BARRA-C2 | AUS-11, AUS-22, AUST-04 | hurs, huss, twiso, twpse | All | Minor | On rare occasions specific humidity fields have unphysical negative, zero, or very small positive value (smaller than 10-6). As well as being unphysical these values cause the wet-bulb temperature calculations to fail at those grid points. | Adjust specific humidity calculations so that there is a minimum value of 2.38E-7. | Fix applied to all data from 2024-12 and onwards. 1979 to 2024 twiso & twpse data fixed in v20250528. 1979 to 2024 huss and hurs data not corrected. |