Bias reduction in decadal predictions of West African monsoon rainfall using regional climate models

Paxian, Andreas; Sein, Dmitry V.; Panitz, Hans-Jürgen; Warscher, Michael; Breil, Marcus; Engel, T.; Tödter, Julian; Krause, Andreas; Cabos, William; Fink, Andreas H.; Ahrens, Bodo; Kunstmann, Harald; Jacob, Daniela; Paeth, Heiko

doi:10.1002/2015jd024143

Cited by 31 publications

(15 citation statements)

References 90 publications

(162 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Similar results were obtained by Hansingo and Reason (2009) using a global circulation model, which produces anomalous precipitation along the Angolan and northern Namibian coast by imposing increased Atlantic SST offshore Angola in the forcing data. Furthermore, this work also coincides with the results of Paxian et al (2016), wherein West African rainfall predictions could be improved by reducing the tropical Atlantic SST bias when regional predictions were coupled to a global ocean model. We have shown that the usage of a regional climate model with higher resolved atmospheric processes in combination with corrected SSTs lead to more realistic representation of the atmospheric water cycle similar to results of Paeth et al (2017).…”

Section: Discussionsupporting

confidence: 85%

“…These inconsistencies are strongest in the eight outermost gridboxes of the model domain, which are used as lateral forcing in REMO, and weakens within the model domain in which the atmosphere can react freely to the SST. Thirdly, the achieved improvements in simulating the atmospheric water cycle using a corrected Atlantic and Indian SSTs are only valid for this region and experiment setup, and may differ in other experiments configurations as found by Paxian et al (2016). Finally, in this work only one regional climate model was applied for downscaling five years global model forcing data.…”

Section: Discussionmentioning

confidence: 88%

See 1 more Smart Citation

Sensitivity of the atmospheric water cycle to corrections of the sea surface temperature bias over southern Africa in a regional climate model

2017

View full text Add to dashboard Cite

Regional climate models (RCMs) have been used to dynamically downscale global climate projections at high spatial and temporal resolution in order to analyse the atmospheric water cycle. In southern Africa, precipitation pattern were strongly affected by the moisture transport from the southeast Atlantic and southwest Indian Ocean and, consequently, by their sea surface temperatures (SSTs). However, global ocean models often have deficiencies in resolving regional to local scale ocean currents, e.g. in ocean areas offshore the South African continent. By downscaling global climate projections using RCMs, the biased SSTs from the global forcing data were introduced to the RCMs and affected the results of regional climate projections. In this work, the impact of the SST bias correction on precipitation, evaporation and moisture transport were analysed over southern Africa. For this analysis, several experiments were conducted with the regional climate model REMO using corrected and uncorrected SSTs. In these experiments, a global MPI-ESM-LR historical simulation was downscaled with the regional climate model REMO to a high spatial resolution of 50 × 50 km 2 and of 25 × 25 km 2 for southern Africa using a double-nesting method. The results showed a distinct impact of the corrected SST on the moisture transport, the meridional vertical circulation and on the precipitation pattern in southern Africa. Furthermore, it was found that the experiment with the corrected SST led to a reduction of the wet bias over southern Africa and to a better agreement with observations as without SST bias corrections.

show abstract

Section: Discussionsupporting

confidence: 85%

Section: Discussionmentioning

confidence: 88%

Sensitivity of the atmospheric water cycle to corrections of the sea surface temperature bias over southern Africa in a regional climate model

2017

View full text Add to dashboard Cite

show abstract

“…Also, both simulations underestimate the latent heat fluxes in the Coast of Guinea, the transitional zone, and over the Sahel. Paxian et al [] and Panitz et al [] relate this uncertainty in CCLM simulations to an impact of land surface conditions and to a misreproduction of the soil water content. Land use and soil characteristics influence the soil moisture that plays a key roll in the partitioning between the surface sensible and latent energy fluxes [ Guillod et al , ].…”

Section: Resultsmentioning

confidence: 99%

Evaluation of the COSMO‐CLM high‐resolution climate simulations over West Africa

et al. 2017

View full text Add to dashboard Cite

The evaluation of a high‐resolution simulation at 0.11° (12 km) with the COnsortium for Small‐scale MOdelling in CLimate Mode (CCLM) regional climate model, applied over West Africa, is presented. This simulation is nested in a CCLM run at resolution of 0.44°, driven with boundary forcing data from the ERA‐Interim reanalysis, and covers the period from 1981 to 2010. The simulated temperature and precipitation are evaluated using three selected climate indices for temperature and eight indices for precipitation in five different regions against a new, daily precipitation climatology covering West Africa and against other state of the art climatologies. The obtained results indicate that CCLM is able to reproduce the observed major climate characteristics including the West African Monsoon within the range of comparable regional climate modeling evaluations studies, but substantial uncertainties remain, especially in the Sahel zone. The seasonal mean temperature bias for the rainy season from June to September ranges from −0.8°C to −1.1°C. The CCLM simulations also underestimate the observed precipitation with biases in precipitation reaching −10% in the high‐resolution and −20% in the low‐resolution model runs. CCLM extends the monsoon precipitation belt too far north, which results in an overestimation of precipitation in the Sahel zone of up to 60%. In the coastal zone, the precipitation is underestimated by up to −90%. These biases in precipitation amounts are associated with errors in the precipitation seasonality. The added value of the higher resolution of the nested run is reflected in a smaller bias in extreme precipitation statistics with respect to the reference data.

show abstract

“…It comprises 40 z-coordinate vertical levels. These simulations were used in a number of evaluation studies regarding decadal predictions of West African monsoon rainfall (Paxian et al 2016;Paeth et al 2017) and the impact of South Atlantic Anticyclone in the generation of coupled model biases in the Angola-Benguela frontal zone (Cabos et al 2017).…”

Section: Regional Climate Simulationsmentioning

confidence: 99%

Assessing the climate change impact on the North African offshore surface wind and coastal low-level jet using coupled and uncoupled regional climate simulations

et al. 2018

View full text Add to dashboard Cite

The North African Coastal Low-Level Jet (NACLLJ) is a semi-permanent feature offshore the north western African coast, linked to the cold nearshore upwelling of the Canary Eastern Boundary Current system. Its main synoptic drivers are the Azores Anticyclone over the ocean and the inland Sahara thermal low. The coastal jet events occur in one of the world's most productive fisheries region, thus the evaluation of the effects of global warming in its properties is imperative. This study proposes an analysis of the annual and intra-annual attributes of the NACLLJ for two time periods 1976-2005 (historical) and 2070-2199 (future), resorting to coupled and uncoupled atmosphere-ocean simulations with the ROM model, as well as near surface offshore wind speed from the CORDEX-Africa ensemble. The future simulations follow the RCP8.5 greenhouse gas emissions scenario. Overall, the ROM coupled simulation presents the best performance in reproducing the present-climate near surface wind speed, offshore northwest Africa, compared to the remaining RCM simulations. The higher SST resolution in the coupled simulations favours much localised colder upwelling strips near the coast and consequently stronger jets. In future climate, a small increase in the surface wind speed is projected, mainly linked to the regions of coastal jet presence. The NACLLJ is projected to be more frequent and intense, encompassing larger areas. An increase of the jet seasonal frequencies of occurrence is projected for all seasons, which is larger from spring to autumn (up to 15, 16 and 22% more frequent, respectively). However, in some offshore areas the winter NACLLJ persistency is likely to double, relatively to present-climate. Higher inter-annual variability is also projected for the future NACLLJ seasonal frequencies. The strengthening of the coastal jet speeds is also significant, between 5 and 12% in all seasons. Additionally, the jet's diurnal cycle shows an increase in jet occurrence across the day, particularly in the mid and late afternoon.

show abstract

Bias reduction in decadal predictions of West African monsoon rainfall using regional climate models

Cited by 31 publications

References 90 publications

Sensitivity of the atmospheric water cycle to corrections of the sea surface temperature bias over southern Africa in a regional climate model

Sensitivity of the atmospheric water cycle to corrections of the sea surface temperature bias over southern Africa in a regional climate model

Evaluation of the COSMO‐CLM high‐resolution climate simulations over West Africa

Assessing the climate change impact on the North African offshore surface wind and coastal low-level jet using coupled and uncoupled regional climate simulations

Contact Info

Product

Resources

About