Progress report on PROMISE from CNRM (18-th month)

WP1000:

A set of statistical procedures in the PV-Wave language has been assembled for the analysis of intraseasonal variability in climate simulations.  They include the spectral methods for the analysis of time series (methods of Hayashi, wavelets) and standard and complex EOF methods for the analysis of the spatial patterns of variability.  These procedures are now being converted into a set of Fortran programs with graphical display using the GRADS software so as to be useable on a wider range of computers.

WP1300:

Several sets of global atmospheric experiments based on ARPEGE version 2 were performed in order to highlight the relevance of soil moisture boundary conditions for simulating the northern hemisphere summer climate and its interannual variability.  The soil moisture influence was demonstrated through the use of a global or regional relaxation of the total soil water content towards the GSWP (Global Soil Wetness Project) monthly climatology covering the 1987-1988 period.  Ensembles of 6-month simulations were performed for both March to September 1987 and 1988, using either free-running or relaxed soil moisture, showing that realistic soil moisture boundary conditions were necessary for simulating the mid-latitude stationary waves, with a significant impact on both extratropical and tropical precipitation (Douville and Chauvin, 2000; Douville et al.  2001; Douville 2001).

In order to test the robustness of this finding, 4 additional sets of experiments have been recently launched over a 15-year period and with the version 3 (semi-lagrangian advection scheme and improved physical package) of the ARPEGE-Climat model.  Since the focus is mainly on the Asian and African monsoon climate, each ensemble is made up of ten 4-month simulations for each of the 15 monsoon seasons (June to September).  In the 1st set, the land surface hydrology is fully interactive and soil moisture is running free.  In the 2nd set, the total soil water content is relaxed towards the monthly climatology derived from the 1st ensemble, so that the interannual variability of soil moisture is strongly damped.  In the 3rd set, the total soil water content is relaxed towards the GSWP climatology:  the soil moisture climatology is more realistic but there is still no interannual variability.  In the 4th and last set, the monthly anomalies derived from the 1st set are superimposed onto the GSWP climatology.  These 4 experiments have been recently completed and will be now analysed using the ANOVA technique.

WP2000:

The transient future climate simulation for the 21-st century with the coupled version of ARPEGE- Climate has been analysed further, concentrating on the African monsoon and the hydrological cycle.  It consists of two 150-year simulations starting in 1950:  a control simulation (SC2) in which the greenhouse gas concentrations and aerosols are kept fixed at their 1950 observed value, and a scenario simulation (SGO) in which the greenhouse gas concentrations are changed annually according to the new scenario SRES-B2 of IPCC.  The climate response over Africa has been analysed in detail for the periods 1980-2010 and 2070-2100.  The climate of the recent period has been validated by comparison with ECMWF analyses and shows that the coupled model is able to reproduce a realistic seasonal cycle of the African monsoon with however a tendency to give a somewhat stronger monsoon than observed.  A detailed analysis of the hydrological cycle and of moisture transport has shown that the increase of low-level moisture convergence is mainly due to the increased water content in a warmer atmosphere and to a northward displacement of the convergence zone over Africa.  A paper on these results has been submitted to Climate Dynamics (Maynard et al, 2001)

The simulated change on the global hydrological cycle has also been analysed in detail, especially the difference between two 30-year time-slices (1970-2000 and 2070-2100 respectively).  The model achieves a reasonable simulation of the present-day precipitation climatology, especially in the Indian and African monsoon regions.  The precipitation changes simulated throughout the 21st century show a stronger convection over the ITCZ, a relative drying in the subtropics, and enhanced monsoon precipitation over India and Africa.  The analysis of the atmospheric branch of the water cycle highlights some of the parameters that control these precipitation anomalies:  changes in precipitable water, water vapor recycling, moisture convergence and precipitation efficiency.  In order to reduce the spread between climate scenarios, more efforts should be devoted to validate these parameters in present-day climate simulations and to analyse their possible trend in observational records. The terrestrial hydrological impacts of global warming have been also investigated, but are probably even more uncertain than the precipitation changes.  This is due to the sensitivity of the simulated surface water balance to deficiencies in both the atmospheric and land surface models. River discharges derived from present-day simulated runoffs still show large errors and are not accurate enough to quantify the response of river basins to increasing amounts of greenhouse gases.  Providing regional hydrological scenarios remains a challenge for the global climate modeling community and downscaling techniques are still necessary for this purpose.

WP 2100:

Using the SSTs from the scenario simulation, as boundary forcing, "time-slice" integrations have been performed with the variable resolution version of Arpege-Climat using its zooming technique to achieve a resolution of about 100 km over Africa.  In order to improve the geographical distribution of the vegetation and of its properties, a new database for vegetation (Ecoclimats) has been used in these experiments.  Two 10-year simulations have been performed using the improved current vegetation distribution:  a control simulation for the current climate using climatological SSTs over 1970-1990, and a future climate simulation for the period 2040-2060 using the SST anomalies from the climate scenario.  The period 2040-2060 is being analysed to show the improvements that the correction of the SST bias and the use of a higher resolution over Africa can bring to the regional simulation of the African monsoon response.  A third simulation taking into account expected changes in land cover over this period as computed by the integrated impact assessment model IMAGE 2.2 in scenario SRES-B2 is in preparation.

References:

Douville H, Chauvin F (2000):  Relevance of soil moisture for seasonal climate predictions:  a preliminary study.  Climate Dynamics, 16:719-736.

Douville H, Chauvin F, Broqua H (2001):  Influence of soil moisture on the Asian and African monsoons.  Part I:  Mean monsoon and daily precipitation.  J Climate, 14:2381-2403.

Douville H (2001):  Influence of soil moisture on the Asian and African monsoons.  Part II:  interannual variability.  J Climate (accepted)

Douville H, Chauvin F, Royer J-F, Salas-MŽlia D, Tyteca S (2001): Sensitivity of global and regional hydrological cycles to increasing amounts of greenhouse gases and aerosols.  (in preparation)

Maynard K, Royer JF, Chauvin F (2001):  Impact of greenhouse warming on the West African Monsoon.  Note de Travail du GMGEC No.  77(Juin), 36 pp.((Submitted to Climate Dynamics))