REGIONAL CLIMATE SIMULATIONS WITH MOISTURE TRACERS TO
INVESTIGATE THE TERRESTRIAL WATER CYCLE OVER THE IBERIAN PENINSULA
Gonzalo Miguez-Macho (1), Alexandre Rios-Entenza (1) and Francina Dominguez (2)
(1) Non-linear Physics Group, Universidade de Santiago de Compostela, Galicia, Spain (email: [email protected])
(2) Department of Atmospheric Sciences, University of Arizona, Tucson, AZ, USA
• Simulations for 3 month cases :May of 2002, 2004 and 2008
SUMMARY
• May 2002: a winter like pattern with strong westerlies and
Atlantic storms and fronts reaching the Iberian Peninsula.
In the semiarid interior of the Iberian Peninsula, the topographic insulation from the surrounding seas promotes
the role of internal sources of moisture and water recycling in the rainfall regime. Inland Iberia, the annual cycle of
precipitation often has a distinctive peak in the springtime, when evapotranspiration is the highest, in contrast to the
• May 2004: Atlantic influence in the first week, but blocking
pattern with easterly Mediterranean flow for the rest.
coastal areas, where it follows more external moisture availability and synoptic forcing, with a maximum in winter-autumn
and a pronounced minimum in the summer. We present here regional climate simulations using the WRF model, with the
new added capability of moisture tracers as a tool to explicitly investigate the sources and pathways of humidity in the
• May 2008 : record high precipitation. Blocking high up north; the
south branch of the split jet over Iberia; cut of lows lingering
around all month long with strong convection and moisture flux
from the Atlantic.
terrestrial water cycle over the Iberian Peninsula. Six new prognostic variables were added in the WRF model
corresponding to tracers for water vapor and other moisture species (cloud water, ice, snow, rain and graupel) originating
from moisture evapotranspired in the selected region (the land mass of the Iberian Peninsula). The new moisture
variables are advected, diffused and convectively and turbulently mixed identically to the full moisture species including
all humidity sources (terrestrial, maritime and lateral boundaries). In addition, the convective scheme (Kein-Fritsch) and
the microphysics scheme (WSM6) are modified to account for generation and depletion of the different tracer species,
IN = 25.38 mm/day
assuming that within one model layer in a grid cell moisture of all origins is well mixed. With these modifications we are
IN = 22.8 mm/day
ρ = 8.6 %
ET = 2.39 mm/day
IN = 27.0 mm/day
ρ = 9.8 %
ET = 2.49 mm/day
ρ = 8.6 %
ET = 2.53 mm/day
able to explicitly track the cycle of water of terrestrial origin and exactly measure its contribution to rainfall with high
resolution within our region of interest. We contrast our results with diagnostic estimates of the recycling ratio
Brubaker’s method
(Brubaker et al., J. of Clim., 1993)
Assumes the recycling ratio ρ
is homogeneous
ρ=
IN
IN + ET
Moisture influx
following the method of Brubaker and of Eltahir and Bras.
Eltahir & Bras method
€
Motivation and goals: spring water cycle in the Iberian Peninsula
The recycling ratio ρ varies spatially but not
temporally; uses mean moisture flux to calculate
• The precipitation cycle in the interior of the
Iberian Peninsula is distinct from that of the
Atlantic and the Mediterranean coastal areas.
• It has a maximum in the spring (May),
secondary of principal toward the E. and NE.
(Eltahir and Bras, QJRM, 1994)
ρ = 13.4 %
ρ = 19.6 %
ρ = 11.9 %
• We investigate the role of ET fluxes and
local control on precipitation dynamics there.
With moisture tracers (parent grid:
Kein-Fritsch convection)
Fig. 1 (a) Monthly mean annual evolution of precipitation for 28 Iberian stations (mm) for the period 1971-2000 and (c) Position of
May in the climatological (1950-2002) monthly ranking of precipitation: May sees the absolute maximum of precipitation in large
areas in the East and North-East. Source: Herrera et al., Int J Climatol , 2010, Belo-Pereira et al. JGR, 2011.
More recycling in convective regimes with weaker external moisture supply
• We perform simulations with the WRF
model with 2 nested grids (20km and 5km
resolutions) for monthly periods.
2002
2004
2008
• Convection is parameterized in the parent
domain and resolved in the nested one.
ρ = 8.7 %
ρ = 13.7 %
ρ = 12.7 %
Fig. 2 Model parent and nested domains with topography (m).
Fig 4. Daily time series of precipitation (blue bars, cm, scale on the left), convective precipitation (outlined bars, cm, scale on the left),
fraction of the total precipitation that is convective (green line, scale on the left) and the recycling ratio (black line, scale on the right).
With moisture tracers (nested grid:
resolved convection)
Water Recycling: the recycling ratio ρ
Same as above: parameterized convection or resolved, recycling is
stronger in convective regimes and weaker with frontal precipitation
• Water recycling refers to the contribution of local ET fluxes to precipitation.
• The recycling ratio ρ is the fraction of precipitation coming from ET. It can be
regarded as an indication of the degree of control of local processes on
precipitation dynamics in a region.
2002
2004
2008
• We test two analytical methods to compute ρ: Brabaker’s and Eltahir and Bras’
(from Schär et al., 1999)
• We compare the results to a explicit calculation using the total precipitation and
the precipitation from moisture tracers from within the selected region.
ρ = 6.3 %
ρ = 10.4 %
Fig 5. Daily time series of precipitation (blue bars, cm, scale on the left) and the recycling ratio (black line, scale on the right).
CONCLUSIONS
Moisture tracers: tracking evapotranspired water from a selected region
Water vapor tracer mixing ratio (g/kg) σ=1
Water vapor tracer mixing ratio (g/kg) σ=10
Fig. 3. Mixing ratio (g/kg) of tracer water vapor (water vapor originating from ET in the Iberian Peninsula) at model σ level 1 (above the surface ) and σ=10 (above the
boundary layer) for 29 May 2004 at 18Z. Arrows are wind vectors at the same levels. The tracer concentration is not well mixed in the column when there is wind shear.
• The recycling ratio varies spatially and temporally,
depending on synoptics and location.
• We introduce 6 extra moisture
species (water vapor, cloud water, ice,
snow, rain and graupel) to trace water
originating from ET in the Iberian
Peninsula.
• The tracer variables do not interfere
with their full moisture counterparts;
they are advected, turbulently mixed
with the same strength, go through
microphysics and convective mixing
assuming that moisture from all origins
is well mixed within each level above
each grid cell.
ρ = 9.8 %
• More recycling downwind, with convective regimes and
weaker external moisture suply.
Fig. 6. Representation of the water fluxes in the Iberian hydrological cycle for
the selected months. Precipitation for each month is taken as 100 units.
Moisture tracers allow us to obtain a
more accurate picture of the water
cycle
• Brubaker’s or Eltahir&Bras’ recycling ratios are not
accurate when there is spatio-temporal variability in the
regional precipitation regime.
• Moisture tracers allow us to obtain a better picture of
the regional water cycle.