Retrieval of the surface characteristics controlling mineral dust emissions: a focus on the aeolian roughness length Béatrice Marticorena L aboratoire I nteruniversitaire des S ystèmes A tmosphériques (LISA) UMR CNRS 7583, Universités Paris 7 et 12; Créteil, France

Erosion threshold Emission processes Location and periods Emission flux intensity Size-distribution Composition Saltation Sand-blasting Model outputs

 Surface characteristics  Local surface properties Aerodynamic roughness % covered surface  Soil - Size-distribution - Texture - Composition  Meteorological data - wind velocity - soil moisture What are the required input data ? Wind friction velocity Erosion threshold Dust flux

 Surface characteristics

 Local surface properties

Aerodynamic roughness

% covered surface

 Soil

- Size-distribution

- Texture

- Composition

 Meteorological data

- wind velocity

- soil moisture

Why focusing on the aeolian roughness length ? Dust fluxes simulated for a loose dry soil (wind velocity measured in N’Djamena, Chad; 318 days) Z 0 =7.10 -4 cm Z 0 =10 -2 cm Dust events frequency ~ 9 % time (28 events) ~32% time (103 events) - Factor ~ 4

Why focusing on the aeolian roughness length ? Z 0 =7.10 -4 cm Z 0 =10 -2 cm Dust emissions (1° × 1°) 0.52 Mt 0.16 Mt Dust fluxes simulated for a loose dry soil (wind velocity measured in N’Djamena, Chad; 318 days) - Factor ~ 3

How deriving the aeolian roughness length ?   a geometrical/geomorphological approach Hypothese on the shape of the roughness elements Vegetation : half sphere, ellipsoid, .. Pebbles, boulbers: ellipsoid, rectangles … Estimation of the roughness length from the mean height and covering rate

Hypothese on the shape of

the roughness elements

Vegetation : half sphere, ellipsoid, ..

Pebbles, boulbers: ellipsoid, rectangles …

How deriving the aeolian roughness length ?   a geometrical/geomorphological approach High resolution information Geomorphological interpretation  1 × 1° information Callot et al., 2000 - % Surface fraction Soil typology Roughness elements : type, mean height, covering rate

- % Surface fraction

Soil typology

Roughness elements :

type, mean height, covering rate

How deriving the aeolian roughness length ?   a geometrical/geomorphological approach Roughness length of the dominant surface type mapped with a geomorphologic approach Callot et al., 2000

How deriving the aeolian roughness length ?   a geometrical/geomorphological approach LIMITATIONS Quality and quantity of available documentation Direct observations for « calibration »

LIMITATIONS

Quality and quantity of

available documentation

Direct observations for « calibration »

How deriving the aeolian roughness length over arid areas (stable roughness) ?   Remote sensing : empirical approach Correlations between experimentally determined roughness length z 0 and backscatter coefficients  0 AISAR  L band-HV SRL  C band-HV

How deriving the aeolian roughness length over arid areas (stable roughness) ?  Remote sensing : empirical approach × × × × × × ×  Empirical relationship PC = f ( ln(Z 0 )) Mapped aeolian roughness length Composite of the POLDER protrusion coefficient PC = k 1 /k 0 (Marticorena et al., in press) × × × × × × ×

How deriving the aeolian roughness length over arid areas (stable roughness) ?   Remote sensing : methodological approach PC Z 01 , Z 02 , ...., Z 0i  1 ,  2 , ....,  i  01 ,  02 , ....,  0i From local measurements to radar scale From radar to POLDER scale PC {  0 PC  0  0 PC  0

Experimental determination of the surface roughness in the south of Tunisia number and dimensions of the roughness elements Roughness length measurements How deriving the aeolian roughness length over arid areas (stable roughness) ?   Remote sensing : methodological approach 14 Anemometers 4 T° sensors H L ou l Ground

Experimental determination of the surface roughness in the south of Tunisia

number and dimensions of the roughness elements

Roughness length measurements

’ Confirmation of the previous relationship between Z 0 and  How deriving the aeolian roughness length over arid areas (stable roughness) ?   Remote sensing : methodological approach

’ High spatial resolution (~30 m ×30 m) ’ Local to regional scale applications How deriving the aeolian roughness length over arid areas (stable roughness) ?   Remote sensing : methodological approach:

’ Spatial resolution : ~6 km × 6 km ’ Regional to global scale applications Selection of POLDER pixels for which the distribution of  0 is homogeneous How deriving the aeolian roughness length over arid areas (stable roughness) ?   Remote sensing : methodological approach: Linear relation between  0 and PC

What about heterogeneous pixel ? How deriving the aeolian roughness length over arid areas (stable roughness) ?   Remote sensing : methodological approach: Consequences on dust flux and dust event frequency ? Regional scale (South Tunisia) - Simulation of the dust fluxes using high resolution map of Z 0 - Comparison with a low resolution map

1990 1992 Annual precipitation Maximum Roughness length Difference between the number of dusty days with and without annual vegetation Simulation of the annual Sahelian vegetation : STEP model (CESBIO)  mean height and covering rate How deriving the aeolian roughness length over semi-arid areas (annual vegetation) ?   A modelling approach:

LIMITATIONS ’ Aerodynamic properties of vegetation - porosity - flexibility - arrangement ’ Temporal and spatial variability of the simulated vegetation sensitivity to the soil type sensitivity to the spatial resolution How deriving the aeolian roughness length over semi-arid areas (annual vegetation) ?   A modelling approach:

LIMITATIONS

’ Aerodynamic properties of vegetation

- porosity

- flexibility

- arrangement

’ Temporal and spatial variability of the simulated vegetation

sensitivity to the soil type

sensitivity to the spatial resolution

How better deriving the aeolian roughness length ? MAIN ISSUES ’ Arid areas : Test on the sensitivity of the dust emission for heterogeneous surface at the regional scale Evaluation of the heterogeneity of the surfaces ? ’ Semi-arid « vegetated » areas Vegetation aerodynamic properties Soil properties Sensitivity to the spatial and temporal resolution of the input and validation data

MAIN ISSUES

’ Arid areas :

Test on the sensitivity of the dust emission for heterogeneous surface at the regional scale

Evaluation of the heterogeneity of the surfaces ?

’ Semi-arid « vegetated » areas

Vegetation aerodynamic properties

Soil properties

Sensitivity to the spatial and temporal resolution of the input and validation data

THANKS to Program on POLDER retrieval of the roughness length : G. Bergametti*, P . Chazette#, F. Dulac#, B. Laurent*, M. Legrand ¶ , F. Maignan# and C. Schmechtig* * LISA, Universités Paris VII-XII, UMR CNRS 7583, Créteil, France # LSCE, UMR CEA-CNRS 1572, Saclay, France ¶ LOA, UMR CNRS XXXX, Villeneuve d’Ascq, France Program on surface roughness in Southern Tunisia : G. Bergametti * , Y . Callot ¤ , P . Chazette # , M . Kardous°, H . Khatteli°, S . Le Hégarat- Mascle + , M. Maillé * , J.L. Rajot * , D . Vidal Madjar + and M . Zribi + ; * LISA, Universités Paris VII-XII, UMR CNRS 7583, Créteil, France ¤ Maison de L’Orient, FRE 2654 , Lyon, France # LSCE, UMR CEA-CNRS 1572, Saclay, France ° Institut des Régions Arides, Médénine, Tunisie + CETP, UMR CNRS 8639, Velizy, France Program on the dust simulation over the Sahel : M . Auvray * , G. Bergametti*, F. Fécan * and E. Mougin‡ * LISA, Universités Paris VII-XII, UMR CNRS 7583, Créteil, France ‡ CESBIO, UMR 5126 CNRS , Toulouse, France