RegCM 4.1 Regional Climate Model User Manual

RegCM 4.1 Regional Climate Model User Manual

The RegCM 4.1 is a regional climate model developed at the Abdus Salam International Centre for Theoretical Physics. It has been used for a wide range of regional climate studies, from process studies to paleo-climate and future climate projections. RegCM 4.1 is a public, open source, user friendly, and portable code that can be applied to any region of the World.

advertisement

Assistant Bot

Need help? Our chatbot has already read the manual and is ready to assist you. Feel free to ask any questions about the device, but providing details will make the conversation more productive.

Regional Climate Model RegCM 4.1 User Manual | Manualzz
The Abdus Salam
International Centre for Theoretical Physics
Strada Costiera, 11 I - 34151 Trieste, Italy
Earth System Physics Section - ESP
Regional Climatic Model RegCM User Manual
Version 4.1
Nellie Elguindi, Xunqiang Bi, Filippo Giorgi, Badrinath Nagarajan,
Jeremy Pal, Fabien Solmon, Sara Rauscher, Ashraf Zakey and Graziano Giuliani
Trieste, Italy
May 2011
2
Acknowledgements
This paper is dedicated to those that have contributed to the growth of RegCM system over the past 20+ years, the
members (800+) of the RegCNET, and the ICTP.
3
4
Contents
1 The REGional Climate Model (RegCM)
9
2 Description
2.1 History . . . . . . . . . . . . . . . . . . . . . .
2.2 Model components . . . . . . . . . . . . . . . .
2.3 The RegCM Model Horizontal and Vertical Grid
2.4 Map Projections and Map-Scale Factors . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
10
10
11
12
13
3 Model Physics
3.1 Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.1 Horizontal Momentum Equations . . . . . . . . . . . .
˙ Equations . . . . . . . . .
3.1.2 Continuity and Sigmadot (σ)
3.1.3 Thermodynamic Equation and Equation for Omega (ω)
3.1.4 Hydrostatic Equation . . . . . . . . . . . . . . . . . . .
3.2 Physics parametrizations . . . . . . . . . . . . . . . . . . . . .
3.2.1 Radiation Scheme . . . . . . . . . . . . . . . . . . . .
3.2.2 Land Surface Models . . . . . . . . . . . . . . . . . . .
3.2.3 Planetary Boundary Layer Scheme . . . . . . . . . . . .
3.2.4 Convective Precipitation Schemes . . . . . . . . . . . .
3.2.5 Large-Scale Precipitation Scheme . . . . . . . . . . . .
3.2.6 Ocean flux Parameterization . . . . . . . . . . . . . . .
3.2.7 Prognostic Sea Surface Skin Temperature Scheme . . .
3.2.8 Pressure Gradient Scheme . . . . . . . . . . . . . . . .
3.2.9 Lake Model . . . . . . . . . . . . . . . . . . . . . . . .
3.2.10 Aerosols and Dust (Chemistry Model) . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
15
15
15
15
16
16
17
17
17
20
21
23
24
24
24
25
25
4 Future Developments
4.1 UFW PBL scheme . . . . . .
4.2 Tiedtke convection scheme . .
4.3 Chemistry . . . . . . . . . . .
4.4 Coupling . . . . . . . . . . .
4.5 2D parallelization . . . . . . .
4.6 Parallel I/O . . . . . . . . . .
4.7 Semi-Lagrangian dynamic core
4.8 Non-Hydrostatic core . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
26
26
26
26
26
27
27
27
27
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6
List of Figures
2.1
2.2
Schematic representation of the vertical structure of the model. This example is for 16 vertical
layers. Dashed lines denote half-sigma levels, solid lines denote full-sigma levels. (Adapted from
the PSU/NCAR Mesoscale Modeling System Tutorial Class Notes and User’s Guide.) . . . . . . .
Schematic representation showing the horizontal Arakawa B-grid staggering of the dot and cross
grid points. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7
12
13
List of Tables
3.1
3.2
3.3
Land Cover/Vegetation classes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
BATS vegetation/land-cover . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Resolution for CLM input parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8
18
19
20
Chapter 1
The RegCM
The RegCM is a regional climate model developed throughout the years, with a wide base of model users. It has
evolved from the first version developed in the late eighties (RegCM1, Dickinson et al. [1989]), Giorgi [1990]), to
later versions in the early nineties (RegCM2, Giorgi et al. [1993b], Giorgi et al. [1993c]), late nineties (RegCM2.5,
Giorgi and Mearns [1999]) and 2000s (RegCM3, Pal et al. [2000]).
The RegCM has been the first limited area model developed for long term regional climate simulation, it has
participated to numerous regional model intercomparison projects, and it has been applied by a large community
for a wide range of regional climate studies, from process studies to paleo-climate and future climate projections
(Giorgi and Mearns [1999], Giorgi et al. [2006]).
The RegCM system is a community model, and in particular it is designed for use by a varied community
composed by scientists in industrialized countries as well as developing nations (Pal et al. [2007]).
As such, it is designed to be a public, open source, user friendly and portable code that can be applied to any
region of the World. It is supported through the Regional Climate research NETwork, or RegCNET, a widespread
network of scientists coordinated by the Earth System Physics section of the Abdus Salam International Centre for
Theoretical Physics Abdus Salam International Centre for Theoretical Physics (ICTP), being the foster the growth
of advanced studies and research in developing countries one of the main aims of the ICTP.
The home of the model is:
http://users.ictp.it/RegCNET
Scientists across this network (currently subscribed by over 750 participants) can communicate through an
email list and via regular scientific workshops, and they have been essential for the evaluation and sequential
improvements of the model.
Since the release of RegCM3 described by Pal et al. [2007], the model has undergone a substantial evolution
both in terms of software code and physics representations, and this has lead to the development of a fourth version
of the model, RegCM4, which was released by the ICTP in June 2010 as a prototype version (RegCM4.0) and in
May 2011 as a first complete version (RegCM4.1).
The purpose of this Manual is to provide a basic reference for RegCM4, with a description of the model, with
a special accent to the improvements recently introduced. Compared to previous versions, RegCM4 includes
new land surface, planetary boundary layer and air-sea flux schemes, a mixed convection and tropical band
configuration, modifications to the pre-existing radiative transfer and boundary layer schemes and a full upgrade
of the model code towards improved flexibility, portability and user friendliness.
The model can be interactively coupled to a 1D lake model, a simplified aerosol scheme (including OC, BC,
SO4, dust and sea spray) and a gas phase chemistry module (CBM-Z). Overall, RegCM4 shows an improved
performance in several respects compared to previous versions, although further testing by the user community is
needed to fully explore its sensitivities and range of applications.
The RegCM is available on the World Wide Web thanks to the Democritos Italy CNR group at:
https://eforge.escience-lab.org/gf/project/regcm
9
Chapter 2
Description
2.1 History
The idea that limited area models (LAMs) could be used for regional studies was originally proposed by Dickinson
et al. [1989] and Giorgi [1990].
This idea was based on the concept of one-way nesting, in which large scale meteorological fields from General
Circulation Model (GCM) runs provide initial and time-dependent meteorological lateral boundary conditions
(LBCs) for high resolution Regional Climate Model (RCM) simulations, with no feedback from the RCM to the
driving GCM.
The first generation NCAR RegCM was built upon the National Center for Atmospheric Research (NCAR)Pennsylvania State University (PSU) Mesoscale Model version 4 (MM4) in the late 1980s [Dickinson et al., 1989;
Giorgi, 1989]. The dynamical component of the model originated from the MM4, which is a compressible, finite
difference model with hydrostatic balance and vertical σ-coordinates.
Later, the use of a split-explicit time integration scheme was added along with an algorithm for reducing
horizontal diffusion in the presence of steep topographical gradients [Giorgi et al., 1993a, b].
As a result, the dynamical core of the RegCM is similar to that of the hydrostatic version of Mesoscale Model
version 5 (MM5) [Grell et al., 1994]: the RegCM4 is thus a hydrostatic, compressible, sigma-p vertical coordinate
model run on an Arakawa B-grid in which wind and thermodynamical variables are horizontally staggered using a
time-splitting explicit integration scheme in which the two fastest gravity modes are first separated from the model
solution and then integrated with smaller time steps.
For application of the MM4 to climate studies, a number of physics parameterizations were replaced, mostly
in the areas of radiative transfer and land surface physics, which led to the first generation RegCM [Dickinson
et al., 1989; Giorgi, 1990]. The first generation RegCM included the Biosphere-Atmosphere Transfer Scheme,
BATS, [Dickinson et al., 1986] for surface process representation, the radiative transfer scheme of the Community
Climate Model version 1 (CCM1), a medium resolution local planetary boundary layer scheme, the Kuo-type
cumulus convection scheme of [Anthes, 1977] and the explicit moisture scheme of [Hsie et al., 1984].
A first major upgrade of the model physics and numerical schemes was documented by [Giorgi et al., 1993a, b],
and resulted in a second generation RegCM, hereafter referred to as REGional Climate Model version 2 (RegCM2).
The physics of RegCM2 was based on that of the NCAR Community Climate Model version 2 (CCM2) [Hack
et al., 1993], and the mesoscale model MM5 [Grell et al., 1994]. In particular, the CCM2 radiative transfer
package [Briegleb, 1992] was used for radiation calculations, the non local boundary layer scheme of [Holtslag
et al., 1990] replaced the older local scheme, the mass flux cumulus cloud scheme of [Grell, 1993] was added as
an option, and the latest version of BATS1E [Dickinson et al., 1993] was included in the model.
In the last few years, some new physics schemes have become available for use in the RegCM, mostly based
on physics schemes of the latest version of the Community Climate Model (CCM), Community Climate Model
version 3 (CCM3) [Kiehl et al., 1996]. First, the CCM2 radiative transfer package has been replaced by that of
the CCM3. In the CCM2 package, the effects of H2 O, O3 , O2 , CO2 and clouds were accounted for by the model.
Solar radiative transfer was treated with a δ-Eddington approach and cloud radiation depended on three cloud
parameters, the cloud fractional cover, the cloud liquid water content, and the cloud effective droplet radius. The
CCM3 scheme retains the same structure as that of the CCM2, but it includes new features such as the effect of
additional greenhouse gases (NO2 , CH4 , CFCs), atmospheric aerosols, and cloud ice. Scattering and absorption of
10
solar radiation by aerosols are also included based on the aerosol optical properties (Absorption Coefficient and
Single Scattering Albedo).
A simplified explicit moisture scheme Hsie et al. [1984] is included, where only a prognostic equation for cloud
water is used, which accounts for cloud water formation, advection and mixing by turbulence, re-evaporation in
sub-saturated conditions, and conversion into rain via a bulk autoconversion term. Prognosed cloud water variable
is directly used in the cloud radiation calculations, and not diagnosed in terms of the local relative humidity, adding
a very important and far reaching element of interaction between the simulated hydrologic cycle and energy budget
calculations.
The solar spectrum optical properties are based on the cloud liquid water path, which is in turn based on the
cloud liquid water amount prognostically calculated by the model, cloud fractional cover, which is calculated
diagnostically as a function of relative humidity, and effective cloud droplet radius, which is parameterized as a
function of temperature and land sea mask for liquid water and as a function of height for ice phase.
In addition, the scheme diagnostically calculates a fraction of cloud ice as a function of temperature. In the
infrared spectrum the cloud emissivity is calculated as a function of cloud liquid/ice water path and cloud infrared
absorption cross sections depending on effective radii for the liquid and ice phase.
One of the problems in this formulation is that the scheme uses the cloud fractional cover to produce grid box
mean cloud properties which are then treated as if the entire grid box was covered by an effectively thinner cloud
layer. However, because of the non-linear nature of radiative transfer, this approach tends to produce a grayer mean
grid box than if separate cloudy and clear sky fractional fluxes were calculated. By taking advantage of the fact
that the scheme also calculates clear sky fluxes for diagnostic purposes, in iRegCM4 we modified this radiative
cloud representation by first calculating the total cloud cover at a given grid point and then calculating the surface
fluxes separately for the cloudy and clear sky portions of the grid box.
The total cloud cover at a model grid box is given by a value intermediate between that obtained using the
random overlap assumption (which maximizes cloud cover) and that given by the largest cloud cover found in any
single layer of the column overlying the grid box (which implies a full overlap and it is thus is a minimum estimate
of total cloud cover).
This modification thus accounts for the occurrence of fractional clear sky at a given grid box, leading to more
realistic grid-box average surface radiative fluxes in fractional cloudy conditions.
A large-scale cloud and precipitation scheme which accounts for the subgrid-scale variability of clouds [Pal
et al., 2000], parameterizations for ocean surface fluxes [Zeng et al., 1998], and multiple cumulus convection
scheme [Anthes, 1977; Grell, 1993; Emanuel, 1991; Emanuel and Zivkovic-Rothman, 1999] are the same as in
RegCM3, but a new ”mixed scheme” Grell+Emanuel is introduced: it allows the user to select one of the two
schemes in function of the ocean-land mask.
The other main development compared to RegCM3 concerns the aerosol radiative transfer calculations. In
RegCM3 the aerosol radiative forcing was based on three dimensional fields produced by the aerosol model, and
included only scattering and absorption in the shortwave spectrum (see Giorgi et al. [2002]). In RegCM4 we added
the contribution of the infrared spectrum following Solmon et al. [2008].
This is especially important for relatively large dust and sea salt particles and it is calculated by introducing
an aerosol infrared emissivity calculated as a function of aerosol path and absorption cross section estimated from
aerosol size distribution and long wave refractive indices. Long wave diffusion, which could be relevant for larger
dust particles, is not treated as part of this scheme.
The mosaic-type parameterization of subgrid-scale heterogeneity in topography and land use [Giorgi et al.,
2003b] allows finer surface resolution in the Biosphere-Atmosphere Transfer Scheme version 1e (BATS1e).
2.2 Model components
The RegCM modeling system has four components: Terrain, ICBC, RegCM, and Postprocessor. Terrain and ICBC
are the two components of RegCM preprocessor. Terrestrial variables (including elevation, landuse and sea surface
temperature) and three-dimensional isobaric meteorological data are horizontally interpolated from a latitudelongitude mesh to a high-resolution domain on either a Rotated (and Normal) Mercator, Lambert Conformal, or
Polar Stereographic projection. Vertical interpolation from pressure levels to the σ coordinate system of RegCM
is also performed. σ surfaces near the ground closely follow the terrain, and the higher-level σ surfaces tend to
approximate isobaric surfaces.
Since the vertical and horizontal resolution and domain size can vary, the modeling package programs employ
11
Figure 2.1: Schematic representation of the vertical structure of the model. This example is for 16 vertical
layers. Dashed lines denote half-sigma levels, solid lines denote full-sigma levels. (Adapted from the PSU/NCAR
Mesoscale Modeling System Tutorial Class Notes and User’s Guide.)
parameterized dimensions requiring a variable amount of core memory, and the requisite hard-disk storage amount
is varied accordingly.
2.3 The RegCM Model Horizontal and Vertical Grid
It is useful to first introduce the model’s grid configuration. The modeling system usually gets and analyzes its data
on pressure surfaces, but these have to be interpolated to the model’s vertical coordinate before input to the model.
The vertical coordinate is terrain-following (Figure 2.1) meaning that the lower grid levels follow the terrain while
the upper surface is flatter. Intermediate levels progressively flatten as the pressure decreases toward the top of the
model. A dimensionless σ coordinate is used to define the model levels where p is the pressure, pt is a specified
constant top pressure, ps is the surface pressure.
σ=
(p − pt )
(ps − pt )
(2.1)
It can be seen from the equation and Figure 2.1 that σ is zero at the top and one at the surface, and each model
level is defined by a value of σ. The model vertical resolution is defined by a list of values between zero and one
12
Figure 2.2: Schematic representation showing the horizontal Arakawa B-grid staggering of the dot and cross grid
points.
that do not necessarily have to be evenly spaced. Commonly the resolution in the boundary layer is much finer
than above, and the number of levels may vary upon the user demand.
The horizontal grid has an Arakawa-Lamb B-staggering of the velocity variables with respect to the scalar
variables. This is shown in Figure 2.2 where it can be seen that the scalars (T, q, p, etc) are defined at the center of
the grid box, while the eastward (u) and northward (v) velocity components are collocated at the corners. The center
points of grid squares will be referred to as cross points, and the corner points are dot points. Hence horizontal
velocity is defined at dot points. Data is input to the model, the preprocessors do the necessary interpolation to
assure consistency with the grid.
All the above variables are defined in the middle of each model vertical layer, referred to as half-levels and
represented by the dashed lines in Figure 2.1. Vertical velocity is carried at the full levels (solid lines). In defining
the sigma levels it is the full levels that are listed, including levels at σ = 0 and 1. The number of model layers is
therefore always one less than the number of full sigma levels.
The finite differencing in the model is, of course, crucially dependent upon the grid staggering wherever
gradients or averaging are represented terms in the equation.
2.4 Map Projections and Map-Scale Factors
The modeling system has a choice of four map projections. Lambert Conformal is suitable for mid-latitudes, Polar
Stereographic for high latitudes, Normal Mercator for low latitudes, and Rotated Mercator for extra choice. The
13
x and y directions in the model do not correspond to west-east and north-south except for the Normal Mercator
projection, and therefore the observed wind generally has to be rotated to the model grid, and the model u and v
components need to be rotated before comparison with observations. These transformations are accounted for in
the model pre-processors that provide data on the model grid (Please note that model output of u and v components,
raw or postprocessed, should be rotated to a lat/lon grid before comparing to observations). The map scale factor,
m, is defined by
m = (distance on grid) / (actual distance on earth)
and its value is usually close to one, varying with latitude. The projections in the model preserve the shape of
small areas, so that dx=dy everywhere, but the grid length varies across the domain to allow a representation of a
spherical surface on a plane surface. Map-scale factors need to be accounted for in the model equations wherever
horizontal gradients are used.
14
Chapter 3
Model Physics
3.1 Dynamics
The model dynamic equations and numerical discretization are described by Grell et al. [1994].
3.1.1
Horizontal Momentum Equations
∂p∗ uσ˙
∂p∗ uu/m ∂p∗ vu/m
−
+
∂x
∂y
∂σ
RTv
∂p∗ ∂φ
∗
−mp
+ f p∗ v + FH u + FV u,
+
(p∗ + pt /σ) ∂x
∂x
∗
∂p∗ v
∂p∗ vσ˙
∂p uv/m ∂p∗ vv/m
−
= −m2
+
∂t
∂x
∂y
∂σ
RTv
∂p∗ ∂φ
−mp∗
+ f p∗ u + FH v + FV v,
+
(p∗ + pt /σ) ∂y
∂y
∂p∗ u
= −m2
∂t
(3.1)
(3.2)
where u and v are the eastward and northward components of velocity, Tv is virtual temperature, φ is
geopotential height, f is the coriolis parameter, R is the gas constant for dry air, m is the map scale factor for
either the Polar Stereographic, Lambert Conformal, or Mercator map projections, σ˙ = dσ
dt , and FH and FV represent
the effects of horizontal and vertical diffusion, and p∗ = ps − pt .
3.1.2
˙ Equations
Continuity and Sigmadot (σ)
∂p∗
= −m2
∂t
∂p∗ σ˙
∂p∗ u/m ∂p∗ v/m
−
+
.
∂x
∂y
∂σ
(3.3)
The vertical integral of Equation 3.3 is used to compute the temporal variation of the surface pressure in the
model,
∂p∗
= −m2
∂t
Z 1 ∗
∂p u/m
0
∂x
∂p∗ v/m
dσ.
+
∂y
(3.4)
˙
After calculation of the surface-pressure tendency ∂p
∂t , the vertical velocity in sigma coordinates (σ) is
computed at each level in the model from the vertical integral of Equation 3.3.
∗
15
σ˙ = −
1
p∗
Z σ ∗
∂p
∂t
0
+ m2
∂p∗ u/m ∂p∗ v/m
+
∂x
∂y
dσ′
(3.5)
˙ = 0) = 0.
where σ′ is a dummy variable of integration and σ(σ
3.1.3
Thermodynamic Equation and Equation for Omega (ω)
The thermodynamic equation is
∂p∗ T
= −m2
∂t
∂p∗ T σ˙
∂p∗ uT /m ∂p∗ vT /m
−
+
+
∂x
∂y
∂σ
RTv ω
p∗ Q
+
+ FH T + FV T
c pm (σ + Pt /past ) c pm
(3.6)
where c pm is the specific heat for moist air at constant pressure, Q is the diabatic heating, FH T represents the
effect of horizontal diffusion, FV T represents the effect of vertical mixing and dry convective adjustment, and ω is
ω = p∗ σ˙ + σ
d p∗
dt
(3.7)
where,
∂p∗
∂p∗
∂p∗
d p∗
=
+m u
+v
dt
∂t
∂x
∂y
(3.8)
The expression for c pm = c p (1 + 0.8qv ), where c p is the specific heat at constant pressure for dry air and qv is
the mixing ratio of water vapor.
3.1.4
Hydrostatic Equation
The hydrostatic equation is used to compute the geopotential heights from the virtual temperature Tv ,
∂φ
qc + qr −1
= −RTv 1 +
∂ln(σ + pt /p∗ )
1 + qv
(3.9)
where Tv = T (1 + 0.608qv ), qv , qc , and qr are the water vapor, cloud water or ice, and rain water or snow, mixing
ratios.
16
3.2 Physics parametrizations
3.2.1
Radiation Scheme
RegCM4 uses the radiation scheme of the NCAR CCM3, which is described in Kiehl et al. [1996]. Briefly, the
solar component, which accounts for the effect of O3 , H2 O, CO2 , and O2 , follows the δ-Eddington approximation
of Kiehl et al. [1996]. It includes 18 spectral intervals from 0.2 to 5 µm. The cloud scattering and absorption
parameterization follow that of Slingo [1989], whereby the optical properties of the cloud droplets (extinction
optical depth, single scattering albedo, and asymmetry parameter) are expressed in terms of the cloud liquid water
content and an effective droplet radius. When cumulus clouds are formed, the gridpoint fractional cloud cover is
such that the total cover for the column extending from the model-computed cloud-base level to the cloud-top level
(calculated assuming random overlap) is a function of horizontal gridpoint spacing. The thickness of the cloud
layer is assumed to be equal to that of the model layer, and a different cloud water content is specified for middle
and low clouds.
3.2.2
Land Surface Models
BATS (default): BATS is a surface package designed to describe the role of vegetation and interactive soil moisture
in modifying the surface-atmosphere exchanges of momentum, energy, and water vapor (see Dickinson et al. [1993]
for details). The model has a vegetation layer, a snow layer, a surface soil layer, 10 cm thick, or root zone layer,
1-2 m thick, and a third deep soil layer 3 m thick. Prognostic equations are solved for the soil layer temperatures
using a generalization of the force-restore method of Deardoff [1978]. The temperature of the canopy and canopy
foilage is calculated diagnostically via an energy balance formulation including sensible, radiative, and latent heat
fluxes.
The soil hydrology calculations include predictive equations for the water content of the soil layers. These
equations account for precipitation, snowmelt, canopy foiliage drip, evapotranspiration, surface runoff, infiltration
below the root zone, and diffusive exchange of water between soil layers. The soil water movement formulation is
obtained from a fit to results from a high-resolution soil model Dickinson [1984] and the surface runoff rates
are expressed as functions of the precipitation rates and the degree of soil water saturation. Snow depth is
prognostically calculated from snowfall, snowmelt, and sublimation. Precipitation is assumed to fall in the form
of snow if the temperature of the lowest model level is below 271 K.
Sensible heat, water vapor, and momentum fluxes at the surface are calculated using a standard surface drag
coefficient formulation based on surface-layer similarity theory. The drag coefficient depends on the surface
roughness length and on the atmospheric stability in the surface layer. The surface evapotranspiration rates
depend on the availability of soil water. Biosphere-Atmosphere Transfer Scheme (BATS) has 20 vegetation types
(Table 3.2; soil textures ranging from coarse (sand), to intermediate (loam), to fine (clay); and different soil colors
(light to dark) for the soil albedo calculations. These are described in Dickinson et al. [1986].
In the latest release version, additional modifications have been made to BATSin order to account for
the subgrid variability of topography and landcover using a mosaic-type approach [Giorgi et al., 2003a].
Thismodification adopts a regular fine-scale surface subgrid for eachcoarse model grid cell. Meteorological
variables are disaggregatedfrom the coarse grid to the fine grid based on the elevationdifferences. The BATS
calculations are then performed separatelyfor each subgrid cell, and surface fluxes are reaggregated onto thecoarse
grid cell for input to the atmospheric model. This parameterization showed a marked improvement in the
representation ofthe surface hydrological cycle in mountainous regions [Giorgi et al., 2003a]. As a first
augmentation, in REGional Climate Model version 4 (RegCM4) two new land use types were added to BATS to
represent urban and sub-urban environments. Urban development not only modifies the surface albedo and alters
the surface energy balance, but also creates impervious surfaces with large effects on runoff and evapotranspiration.
These effects can be described by modifying relevant properties of the land surface types in the BATS package, such
as maximum vegetation cover, roughness length, albedo, and soil characteristics. For this purpose, we implemented
the parameters proposed in Table 1 of Kueppers et al. [2008].
CLM (optional): The Community Land Model (CLM; Oleson et al. [2008]) is the land surface model developed
by the National Center of Atmospheric Research (NCAR) as part of the Community Climate System Model
(CCSM), described in detail in Collins et al. [2006]. CLM version 3.5 was coupled to RegCM for a more detailed
land surface description option. CLM contains five possible snow layers with an additional representation of trace
snow and ten unevenly spaced soil layers with explicit solutions of temperature, liquid water and ice water in each
layer. To account for land surface complexity within a climate model grid cell, CLM uses a tile or mosaic approach
17
Table 3.1:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
Land Cover/Vegetation classes
Crop/mixed farming
Short grass
Evergreen needleleaf tree
Deciduous needleleaf tree
Deciduous broadleaf tree
Evergreen broadleaf tree
Tall grass
Desert
Tundra
Irrigated Crop
Semi-desert
Ice cap/glacier
Bog or marsh
Inland water
Ocean
Evergreen shrub
Deciduous shrub
Mixed Woodland
Forest/Field mosaic
Water and Land mixture
to capture surface heterogeneity. Each CLM gridcell contains up to four different land cover types (glacier, wetland,
lake, and vegetated), where the vegetated fraction can be further divided into 17 different plant functional types.
Hydrological and energy balance equations are solved for each land cover type and aggregated back to the gridcell
level. A detailed discussion of CLM version 3 implemented in RegCM3 and comparative analysis of land surface
parameterization options is presented in Steiner et al. [2009]. Since CLM was developed for the global scale,
several input files and processes were modified to make it more appropriate for regional simulations, including
(1) the use of high resolution input data, (2) soil moisture initialization, and (3) and an improved treatment of grid
cells along coastlines. For the model input data, CLM requires several time-invariant surface input parameters: soil
color, soil texture, percent cover of each land surface type, leaf and stem area indices, maximum saturation fraction,
and land fraction [Lawrence and Chase, 2007]. Table 3.3 shows the resolution for each input parameter used at the
regional scale in RegCM-CLM compared to resolutions typically used for global simulations. The resolution of
surface input parameters was increased for several parameters to capture surface heterogeneity when interpolating
to the regional climate grid. Similar to Lawrence and Chase [2007], the number of soil colors was extended from 8
to 20 classes to resolve regional variations. The second modification was to update the soil moisture initialization
based on a climatological soil moisture average [Giorgi and Bates, 1989] over the use of constant soil moisture
content throughout the grid generally used for global CLM. By using a climatological average for soil moisture,
model spin-up time is reduced with regards to deeper soil layers. The third modification to the CLM is the inclusion
of a mosaic approach for gridcells that contain both land and ocean surface types. With this approach, a weighted
average of necessary surface variables was calculated for land/ocean gridcells using the land fraction input dataset.
This method provides a better representation of coastlines using the high-resolution land fraction data described in
Table 3.3. For a more detailed description of CLM physics parameterizations see Oleson [2004].
18
Table 3.2: BATS vegetation/land-cover
Parameter
19
Max fractional
vegetation cover
Difference between max
fractional vegetation
cover and cover at 269 K
Roughness length (m)
Displacement height (m)
Min stomatal
resistence (s/m)
Max Leaf Area Index
Min Leaf Area Index
Stem (dead matter
area index)
Inverse square root of
leaf dimension (m−1/2 )
Light sensitivity
factor (m2 W−1 )
Upper soil layer
depth (mm)
Root zone soil
layer depth (mm)
Depth of total
soil (mm)
Soil texture type
Soil color type
Vegetation albedo for
wavelengths < 0.7 µ m
Vegetation albedo for
wavelengths > 0.7 µ m
Land Cover/Vegetation Type
9
10
11
12
1
2
3
4
5
6
7
8
13
14
15
16
17
18
19
20
0.85
0.80
0.80
0.80
0.80
0.90
0.80
0.00
0.60
0.80
0.35
0.00
0.80
0.00
0.00
0.80
0.80
0.80
0.80
0.80
0.6
0.08
0.0
0.1
0.05
0.0
0.1
1.00
9.0
0.3
1.00
9.0
0.5
0.80
0.0
0.3
2.00
18.0
0.0
0.10
0.0
0.2
0.05
0.0
0.6
0.04
0.0
0.1
0.06
0.0
0.0
0.10
0.0
0.4
0.01
0.0
0.0
0.03
0.0
0.0
0.0004
0.0
0.2
0.0004
0.0
0.3
0.10
0.0
0.2
0.10
0.0
0.4
0.80
0.0
0.4
0.3
0.0
0.3
0.0
45
6
0.5
60
2
0.5
80
6
5
80
6
1
120
6
1
60
6
5
60
6
0.5
200
0
0
80
6
0.5
45
6
0.5
150
6
0.5
200
0
0
45
6
0.5
200
0
0
200
0
0
80
6
5
120
6
1
100
6
3
120
6
0.5
120
6
0.5
0.5
4.0
2.0
2.0
2.0
2.0
2.0
0.5
0.5
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
10
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
0.02
0.02
0.06
0.06
0.06
0.06
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.06
0.02
0.02
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
1000
1000
1500
1500
2000
1500
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
2000
2000
2000
3000
6
5
3000
6
3
3000
6
4
3000
6
4
3000
7
4
3000
8
4
3000
6
4
3000
3
1
3000
6
3
3000
6
3
3000
5
2
3000
12
1
3000
6
5
3000
6
5
3000
6
5
3000
6
4
3000
5
3
3000
6
4
3000
6
4
3000
0
0
0.10
0.10
0.05
0.05
0.08
0.04
0.08
0.20
0.10
0.08
0.17
0.80
0.06
0.07
0.07
0.05
0.08
0.06
0.06
0.06
0.30
0.30
0.23
0.23
0.28
0.20
0.30
0.40
0.30
0.28
0.34
0.60
0.18
0.20
0.20
0.23
0.28
0.24
0.18
0.18
Table 3.3: Resolution for CLM input parameters
Input data
Grid Spacing Lon range Lat range
Glacier
0.05◦ x 0.05◦ ±179.975 ±89.975
Lake
0.05◦ x 0.05◦ ±179.975 ±89.975
Wetland
0.05◦ x 0.05◦ ±179.975 ±89.975
Land fraction
0.05◦ x 0.05◦ ±179.975 ±89.975
LAI/SAI
0.5◦ x 0.5◦
±179.75
±89.75
PFT
0.5◦ x 0.5◦
±179.75
±89.75
Soil color
0.05◦ x 0.05◦ ±179.975 ±89.975
Soil texture
0.05◦ x 0.05◦ ±179.975 ±89.975
Max. sat. area
0.5◦ x 0.5◦
±179.75
±89.75
3.2.3
Planetary Boundary Layer Scheme
The planetary boundary layer scheme, developed by Holtslag et al. [1990], is based on a nonlocal diffusion
concept that takes into account countergradient fluxes resulting from large-scale eddies in an unstable, well-mixed
atmosphere. The vertical eddy flux within the PBL is given by
Fc = −Kc
∂C
− γc
∂z
(3.10)
where γc is a “countergradient” transport term describing nonlocal transport due to dry deep convection. The
eddy diffusivity is given by the nonlocal formulation
z2
Kc = kwt z 1 −
h
(3.11)
where k is the von Karman constant; wt is a turbulent convective velocity that depends on the friction velocity,
height, and the Monin–Obhukov length; and h is the PBL height.
The countergradient term for temperature and water vapor is given by
γc = C
φc 0
wt h
(3.12)
where C is a constant equal to 8.5, and φc 0 is the surface temperature or water vapor flux. Equation 3.12 is
applied between the top of the PBL and the top of the surface layer, which is assumed to be equal to 0.1h. Outside
this region and for momentum, γc is assumed to be equal to 0.
For the calculation of the eddy diffusivity and countergradient terms, the PBL height is diagnostically computed
from
h=
Ric r[u(h)2 + v(h)2 ]
(g/θs )[θv (h) − θs ]
(3.13)
where u(h), v(h), and θv are the wind components and the virtual potential temperature at the PBL height, g is
gravity, Ric r is the critical bulk Richardson number, and θs is an appropriate temperature of are near the surface.
Refer to Holtslag et al. [1990] and Holtslag and Boville [1993] for a more detailed description.
Compared to other schemes this formulation tends to produce relatively strong, and often excessive, turbulent
vertical transfer. For example, after extensive testing, we found excessive vertical transfer of moisture in the model
resulting in low moisture amounts near the surface and excessive moisture near the PBL top.
Therefore in order to ameliorate this problem, the countergradient term for water vapor was removed in
RegCM4. Another problem of the Holtslag scheme (at least in our implementation) is an excessive vertical
20
transport of heat, moisture and momentum in very stable conditions, such as during the winter in northern
hemisphere high latitude regions. For example we found that in such conditions the scheme fails to simulate
near surface temperature inversions.
This in turn leads to large warm winter biases (even ¿ 10 degrees) over regions such as Northern Siberia
and Northern Canada. As an ad-hoc fix to address this problem, in RegCM4 we implemented the following
modification to the scheme:
• We first define very stable conditions within the Holtslag parameterization as conditions in which the ratio
of the height from the surface over the Monin-Obhukov length is lower than 0.1.
• When such conditions are found, we set to 0 the eddy diffusivity and counter-gradient terms for all variables.
Preliminary tests showed that this modification reduces the warm bias in high latitude winter conditions and
allows the model to better capture surface inversions. These modifications have thus been incorporated as default
in the RegCM4 code.
3.2.4
Convective Precipitation Schemes
Convective precipitation is computed using one of three schemes: (1) Modified-Kuo scheme Anthes [1977]; (2)
Grell scheme Grell [1993]; and (3) MIT-Emanuel scheme [Emanuel, 1991; Emanuel and Zivkovic-Rothman, 1999].
In addition, the Grell parameterization is implemented using one of two closure assumptions: (1) the Arakawa and
Schubert closure Grell et al. [1994] and (2) the Fritsch and Chappell closure Fritsch and Chappell [1980], hereafter
refered to as AS74 and FC80, respectively.
1. Kuo Scheme: Convective activity in the Kuo scheme is initiated when the moisture convergence M in a
column exceeds a given threshold and the vertical sounding is convectively unstable. A fraction of the
moisture convergence β moistens the column and the rest is converted into rainfall PCU according to the
following relation:
PCU
= M(1 − β).
(3.14)
β is a function of the average relative humidity RH of the sounding as follows:
β
=
2(1 − RH)
1.0
RH ≥ 0.5
otherwise
(3.15)
Note that the moisture convergence term includes only the advective tendencies for water vapor. However,
evapotranspiration from the previous time step is indirectly included in M since it tends to moisten the
lower atmosphere. Hence, as the evapotranspiration increases, more and more of it is converted into rainfall
assuming the column is unstable. The latent heating resulting from condensation is distributed between the
cloud top and bottom by a function that allocates the maximum heating to the upper portion of the cloud
layer. To eliminate numerical point storms, a horizontal diffusion term and a time release constant are
included so that the redistributions of moisture and the latent heat release are not performed instantaneously
[Giorgi and Bates, 1989; Giorgi and Marinucci, 1991].
2. Grell Scheme: The Grell scheme Grell [1993], similar to the AS74 parameterization, considers clouds as
two steady-state circulations: an updraft and a downdraft. No direct mixing occurs between the cloudy air
and the environmental air except at the top and bottom of the circulations. The mass flux is constant with
height and no entrainment or detrainment occurs along the cloud edges. The originating levels of the updraft
and downdraft are given by the levels of maximum and minimum moist static energy, respectively. The Grell
scheme is activated when a lifted parcel attains moist convection. Condensation in the updraft is calculated
by lifting a saturated parcel. The downdraft mass flux (m0 ) depends on the updraft mass flux (mb ) according
to the following relation:
21
m0 =
βI1
mb
I2
(3.16)
where I1 is the normalized updraft condensation, I2 is the normalized downdraft evaporation, and β is the
fraction of updraft condensation that re-evaporates in the downdraft. β depends on the wind shear and
typically varies between 0.3 and 0.5. Rainfall is given by
PCU
= I1 mb (1 − β)
(3.17)
Heating and moistening in the Grell scheme are determined both by the mass fluxes and the detrainment at
the cloud top and bottom. In addition, the cooling effect of moist downdrafts is included.
Due to the simplistic nature of the Grell scheme, several closure assumptions can be adopted. RegCM4’s
earlier version directly implements the quasi-equilibrium assumption of AS74. It assumes that convective
clouds stabilize the environment as fast as non-convective processes destabilize it as follows:
mb =
ABE ′′ − ABE
NA∆t
(3.18)
where ABE is the buoyant energy available for convection, ABE ′′ is the amount of buoyant energy available
for convection in addition to the buoyant energy generated by some of the non-convective processes during
the time interval ∆t, and NA is the rate of change of ABE per unit mb . The difference ABE ′′ − ABE can be
thought of as the rate of destabilization over time ∆t. ABE ′′ is computed from the current fields plus the
future tendencies resulting from the advection of heat and moisture and the dry adiabatic adjustment.
In the latest RegCM4 version, by default, we use a stability based closure assumption, the FC80 type
closure assumption, that is commonly implemented in GCMs and RCMs. In this closure, it is assumed
that convection removes the ABE over a given time scale as follows:
mb =
ABE
NAτ
(3.19)
where τ is the ABE removal time scale.
The fundamental difference between the two assumptions is that the AS74 closure assumption relates the
convective fluxes and rainfall to the tendencies in the state of the atmosphere, while the FC80 closure
assumption relates the convective fluxes to the degree of instability in the atmosphere. Both schemes achieve
a statistical equilibrium between convection and the large-scale processes.
A number of parameters present in the scheme can be used to optimize its performance, and Giorgi et al.
[1993c] discusses a wide range of sensitivity experiments. We found that the parameter to which the scheme
is most sensitive is by and large the fraction of precipitation evaporated in the downdraft (Peff, with values
from 0 to 1), which essentially measures the precipitation efficiency. Larger values of Peff lead to reduced
precipitation.
3. MIT-Emanuel scheme: More detailed descriptions can be found in Emanuel [1991] andEmanuel
and Zivkovic-Rothman [1999]. The scheme assumes that the mixing in clouds ishighly episodic and
inhomogeneous (as opposed to a continuousentraining plume) and considers convective fluxes based on
anidealized model of sub-cloud-scale updrafts and downdrafts.Convection is triggered when the level of
neutral buoyancy is greaterthan the cloud base level. Between these two levels, air is liftedand a fraction
of the condensed moisture forms precipitation while theremaining fraction forms the cloud. The cloud is
assumed to mix withthe air from the environment according to a uniform spectrum ofmixtures that ascend
or descend to their respective levels of neutralbuoyancy. The mixing entrainment and detrainment rates are
functionsof the vertical gradients of buoyancy in clouds. The fraction of thetotal cloud base mass flux that
22
mixes with its environment at eachlevel is proportional to the undiluted buoyancy rate of change withaltitude.
The cloud base upward mass flux is relaxed towards thesub-cloud layer quasi equilibrium.
In addition to a more physical representation of convection, the MIT-Emanuel scheme offers several
advantages compared to theother RegCM4 convection options. For instance, it includes aformulation of
the auto-conversion of cloud water into precipitationinside cumulus clouds, and ice processes are accounted
for by allowingthe auto-conversion threshold water content to be temperaturedependent. Additionally, the
precipitation is added to a single,hydrostatic, unsaturated downdraft that transports heat and water. Lastly,
the MIT-Emanuel scheme considers the transport of passive tracers.
The MIT scheme is the most complex of the three and also includes a number of parameters that can be
used to optimize the model performance in different climate regimes. Differently from the Grell scheme,
however, test experiments did not identify a single parameter to which the model is most sensitive.
A major augmentation in RegCM4 compared to previous versions of the model is the capability of running
different convection schemes over land and ocean, a configuration which we refer to as mixed convection.
Extensive test experiments showed that different schemes have different performance over different regions, and
in particular over land vs. ocean areas.
For example, the MIT scheme tends to produce excessive precipitation over land areas, especially through the
occurrence of very intense individual precipitation events.
In other words, once the scheme is activated, it becomes difficult to decelerate. Conversely, we found that the
Grell scheme tends to produce excessively weak precipitation over tropical oceans.
These preliminary tests suggested that a mixed convection approach by which, for example, the MIT scheme
is used over oceans and the Grell scheme over land, might be the most suitable option to pursue, and therefore this
option was added to the model.
3.2.5
Large-Scale Precipitation Scheme
Subgrid Explicit Moisture Scheme (SUBEX) is used to handle nonconvective clouds and precipitation resolved
by the model. This is one of the new components of the model. SUBEX accounts for the subgrid variability in
clouds by linking the average grid cell relative humidity to the cloud fraction and cloud water following the work
of Sundqvist et al. [1989].
The fraction of the grid cell covered by clouds, FC, is determined by,
FC =
r
RH − RHmin
RHmax − RHmin
(3.20)
where RHmin is the relative humidity threshold at which clouds begin to form, and RHmax is the relative humidity
where FC reaches unity. FC is assumed to be zero when RH is less than RHmin and unity when RH is greater than
RHmax .
Precipitation P forms when the cloud water content exceeds the autoconversion threshold Qth c according to the
following relation:
P = Cppt (Qc /FC − Qc th )FC
(3.21)
where 1/Cppt can be considered the characteristic time for which cloud droplets are converted to raindrops.
The threshold is obtained by scaling the median cloud liquid water content equation according to the following:
Qth c = Cacs 10−0.49+0.013T
(3.22)
where T is temperature in degrees Celsius, and Cacs is the autoconversion scale factor. Precipitation is assumed
to fall instantaneously.
SUBEX also includes simple formulations for raindrop accretion and evaporation. The formulation for the
accretion of cloud droplets by falling rain droplets is based on the work of Beheng [1994] and is as follows:
Pacc = Cacc QPsum
23
(3.23)
where Pacc is the amount of accreted cloud water, Cacc is the accretion rate coefficient, and Psum is the
accumulated precipitation from above falling through the cloud.
Precipitation evaporation is based on the work of Sundqvist et al. [1989] and is as follows
Pevap = Cevap (1 − RH)P1/2 sum
(3.24)
where Pevap is the amount of evaporated precipitation, and Cevap is the rate coefficient. For a more detailed
description of SUBEX and a list of the parameter values refer to Pal et al. [2000].
Traditionally, REGional Climate Model version 3 (RegCM3) has shown a tendency to produce excessive
precipitation, especially at high resolutions, and optimizations of the in-cloud liquid water threshold for the
activation of the autoconversion term Qcth and the rate of sub-cloud evaporation Cevap parameters have proven
effective in ameliorating this problem: greater values of Qth and Cevap lead to decreased precipitation amounts.
3.2.6
Ocean flux Parameterization
BATS uses standard Monin-Obukhov similarity relations to compute the fluxes with no special treatment of
convective and very stable conditions. In addition, the roughness length is set to a constant, i.e. it is not a function
of wind and stability.
The Zeng scheme describes all stability conditions and includes a gustiness velocity to account for the
additional flux induced by boundary layer scale variability. Sensible heat (SH), latent heat (LH), and momentum
(τ) fluxes between the sea surface and lower atmosphere are calculated using the following bulk aerodynamic
algorithms,
τ = ρa u∗ 2 (ux 2 + uy 2 )1/2 /u
(3.25)
SH = −ρaCpa u∗ θ∗
(3.26)
LH = −ρa Le u∗ q∗
(3.27)
where ux and uy are mean wind components, u∗ is the frictional wind velocity, θ∗ is the temperature scaling
parameter, q∗ is the specific humidity scaling parameter, ρa is air density, Cpa is specific heat of air, and Le is the
latent heat of vaporization. For further details on the calculation of these parameters refer to Zeng et al. [1998].
3.2.7
Prognostic Sea Surface Skin Temperature Scheme
By default in RegCM, sea surface temperatures (SST) are prescribed every six hours from temporally interpolated
weekly or monthly SST products. These products, which are produced from satellite retrievals and in situ
measurements, are representative of the mean temperature in the top few meters of the ocean. However, the actual
SST can differ significantly from this mean temperature due to the cool-skin and warm-layer effects described
by Fairall et al. [1996]. To improve the calculation of diurnal fluxes over the ocean, the prognostic SST scheme
described by Zeng [2005] was implemented in RegCM4. The scheme is based on a two-layer one-dimensional
heat transfer model, with the top layer representing the upper few millimeters of the ocean which is cooled by net
longwave radiation loss and surface fluxes. The bottom layer is three meters thick, it is warmer by solar radiation
and exchanges heat with the top layer. This diurnal SST scheme appears to provide significant, although not major,
effects on the model climatology mostly over tropical oceans, for example the Indian ocean, and it is now used as
default in RegCM4.
3.2.8
Pressure Gradient Scheme
Two options are available for calculating the pressure gradient force. The normal way uses the full fields. The
other way is the hydrostatic deduction scheme which makes use of a perturbation temperature. In this scheme,
extra smoothing on the top is done in order to reduce errors related to the PGF calculation.
24
3.2.9
Lake Model
The lake model developed by Hostetler et al. [1993] can be interactively coupled to the atmospheric model. In the
lake model, fluxes of heat, moisture, and momentum are calculated based on meteorological inputs and the lake
surface temperature and albedo. Heat is transferred vertically between lake model layers by eddy and convective
mixing. Ice and snow may cover part or all of the lake surface.
In the lake model, the prognostic equation for temperature is
∂T
∂2 T
= (ke + km ) 2
∂t
∂z
(3.28)
where T is the temperature of the lake layer, and ke and km are the eddy and molecular diffusivities, respectively.
The parameterization of Henderson-Sellers [1986] is used to calculate ke and km is set to a constant value of
39 × 10−7 m2 s−1 except under ice and at the deepest points in the lake.
Sensible and latent heat fluxes from the lake are calculated using the BATS parameterizations Dickinson et al.
[1993]. The bulk aerodynamic formulations for latent heat flux (Fq ) and sensible heat flux (Fs ) are as follows,
Fq = ρaCDVa (qs − qa )
Fs = ρaCpCDVa (Ts − Ta )
(3.29)
(3.30)
where the subscripts s and a refer to surface and air, respectively; ρa is the density of air, Va is the wind speed,
Cp , q is specific humidity, and T is temperature. The momentum drag coefficient, CD , depends on roughness length
and the surface bulk Richardson number.
Under ice-free conditions, the lake surface albedo is calculated as a function of solar zenith angle HendersonSellers [1986]. Longwave radiation emitted from the lake is calculated according to the Stefan-Boltzmann law. The
lake model uses the partial ice cover scheme of Patterson and Hamblin [1988] to represent the different heat and
moisture exchanges between open water and ice surfaces and the atmosphere, and to calculate the surface energy
of lake ice and overlying snow. For further details refer to Hostetler et al. [1993] and Small and Sloan [1999].
3.2.10
Aerosols and Dust (Chemistry Model)
The representation of dust emission processes is a key element in a dust model and depends on the wind conditions,
the soil characteristics and the particle size. Following Laurent et al. [2008] and Alfaro and Gomes [2001], here
the dust emission calculation is based on parameterizations of soil aggregate saltation and sandblasting processes.
The main steps in this calculation are: The specification of soil aggregate size distribution for each model grid
cell, the calculation of a threshold friction velocity leading to erosion and saltation processes, the calculation of the
horizontal saltating soil aggregate mass flux, and finally the calculation of the vertical transportable dust particle
mass flux generated by the saltating aggregates. In relation to the BATS interface, these parameterizations become
effective in the model for cells dominated by desert and semi desert land cover.
25
Chapter 4
Future Developments
We have lot of exciting plans for future model improvements, some of which are in a already mature stage and
under testing, with some published results, whereas others are done only on the paper in a whishlist for next years.
Nevertheless we want to share this with users, to have hints and encourage contributions. Some of the development
results/ideas are listed below, in a ”time to market” order.
4.1 UFW PBL scheme
One of the deficiencies identified in RegCM3 has been the lack of simulation of low level stratus clouds, a problem
clearly tied to the excessive vertical transport in the Holtslag PBL scheme (O’ Brien [2011]). To address this
problem Travis OBrien coupled to the RegCM4 the general turbulence closure parameterization of [Grenier and
Bretherton, 2001; Bretherton et al., 2004], which we refer to as UW-PBL. This is a 1.5 order local, down-gradient
diffusion parameterization in which the velocity scale is based on turbulent kinetic energy (TKE). The TKE is in
turn calculated prognostically from the balance of buoyant production/destruction, shear production, dissipation
vertical transport and horizontal diffusion and advection. The scheme also parameterizes the entrainment process
and its enhancement by evaporation of cloudy air into entrained air. One property of the scheme is the use of a
mixing length formulation based on a 2010 paper by Grisogono (ref?) which allows a more realistic description
of sharp inversions under strongly stable conditions. The UW-PBL has been so far tested within the RegCM4
framework mostly in midlatitude domains, such as the continental US (where it considerably improved the
simulation of low level stratus clouds) and Europe.
This scheme is currently in a SVN branch of the code and will be merged into the main development trunk as
soon as the accompanying paper will be published, and will be available in the next model release.
4.2 Tiedtke convection scheme
Adrian Tompkins is developing an adaptation of the ECHAM5.4 Tiedtke [1989] cumulus convection scheme for
the RegCM model. The code from ECHAM has been ported into RegCM, and extensive testing is planned in the
second half of 2011. This option should be available for next model release.
4.3 Chemistry
Fabien Solmon is developing the coupling of RegCM model with the CBMZ chemical module with the Sillmann
fast solver.
4.4 Coupling
We have resolved to adopt for the RegCM model a standard model coupling engine: the Earth System Modeling
Framework (ESMF). Ufuk Utku Turuncoglu is already adapting model data structures to use the ESMF framework.
26
First target will be to couple the RegCM model to the Regional Oceanic Modeling System (ROMS) oceanic model,
and update the Community Land Surface Model (CLM) to version 4.
4.5 2D parallelization
This long standing limitation of the model in the parallel performances will be faced: we plan to drop altogether
the Serial model version (does exist anymore a single core processor?), clean up model parallel code and perform
a dynamical 2D decomposition of the model domain.
4.6 Parallel I/O
This is another limit of the current model implementation, where all data need to be gathered by the master
processor before being written to disk. Again, if running on a decent cluster, all processors usually have access to
disk resources, and a form of parallel I/O will allow a big performance boost as well as a reduction of some of the
MPI communication data at the expenses of an increase of the requirements for the cluster I/O channel.
4.7 Semi-Lagrangian dynamic core
A semi-Lagrangian advection scheme for the water vapor and advection tracers will allow a different timestep for
the transport schemes, which should result in a performance prize.
4.8 Non-Hydrostatic core
We want to implement the non-hydrostatic core to allow physical downscaling of large scale model simulation
under the 20 kilometers limit of the hydrostatic model.
27
Bibliography
Alfaro, S. C., and L. Gomes, Modeling mineral aerosol production by wind erosion: Emission intensities and
aerosol size distributions in source areas, Journal of Geophysical Research, 106, d16, 2001.
Anthes, R. A., A cumulus parameterization scheme utilizing a one-dimensional cloud model, Mon. Wea. Rev., 105,
270–286, 1977.
Beheng, K. D., A parameterization of warm cloud microphysical conversion processes, Atmos. Res., 33, 193–206,
1994.
Bretherton, C. S., J. McCaa, and H. Grenier, A new parameterization for shallow cumulus convection and its
application to marine subtropical cloud-topped boundary layers. part i: Description and 1d results, Monthly
Weather Review, 132, 864–882, 2004.
Briegleb, B. P., Delta-eddington approximation for solar radiation in the ncar community climate model, J.
Geophys. Res., 97, 7603–7612, 1992.
Collins, W. D., et al., The Community Climate System Model version 3 (CCSM3), Journal of Climate, 19, 2122–
2143, 2006.
Deardoff, J. W., Efficient prediction of ground surface temperature and moisture with inclusion of a layer of
vegetation, J. Geophys. Res., 83, 1889–1903, 1978.
Dickinson, R. E., Climate Processes and Climate Sensitivity, chap. Modeling evapotranspiration processes for
three-dimensional global climate models, pp. 52–72, American Geophysical Union, 1984.
Dickinson, R. E., P. J. Kennedy, A. Henderson-Sellers, and M. Wilson, Biosphere-atmosphere transfer
scheme (bats) for the ncar community climate model, Tech. Rep. NCARE/TN-275+STR, National Center for
Atmospheric Research, 1986.
Dickinson, R. E., R. M. Errico, F. Giorgi, and G. T. Bates, A regional climate model for the western United States,
Climatic Change, 15, 383–422, 1989.
Dickinson, R. E., A. Henderson-Sellers, and P. J. Kennedy, Biosphere-atmosphere transfer scheme (bats) version
1e as coupled to the ncar community climate model, Tech. rep., National Center for Atmospheric Research,
1993.
Emanuel, K. A., A scheme for representing cumulus convection in large-scale models, J. Atmos. Sci., 48(21),
2313–2335, 1991.
Emanuel, K. A., and M. Zivkovic-Rothman, Development and evaluation of a convection scheme for use in climate
models, J. Atmos. Sci., 56, 1766–1782, 1999.
Fairall, C., E. Bradley, J. Godfrey, G. Wick, J. Edson, and G. Young, Cool-skin and warm layer effects on sea
surface temperature, Journal of Geophysical Research, 101, 1295–1308, 1996.
Fritsch, J. M., and C. F. Chappell, Numerical prediction of convectively driven mesoscale pressure systems. part i:
Convective parameterization, J. Atmos. Sci., 37, 1722–1733, 1980.
Giorgi, F., Two-dimensional simulations of possible mesoscale effects of nuclear war fires, J. Geophys. Res., 94,
1127–1144, 1989.
28
Giorgi, F., Simulation of regional climate using a limited area model nested in a general circulation model, J.
Climate, 3, 941–963, 1990.
Giorgi, F., and G. T. Bates, The climatological skill of a regional model over complex terrain, Mon. Wea. Rev., 117,
2325–2347, 1989.
Giorgi, F., and M. R. Marinucci, Validation of a regional atmospheric model over europe: Sensitivity of wintertime
and summertime simulations to selected physics parameterizations and lower boundary conditions, Quart. J.
Roy. Meteor. Soc., 117, 1171–1206, 1991.
Giorgi, F., and L. O. Mearns, Introduction to special section: Regional climate modeling revisited, J. Geophys.
Res., 104, 6335–6352, 1999.
Giorgi, F., G. T. Bates, and S. J. Nieman, The multi-year surface climatology of a regional atmospheric model over
the western united states, J. Climate, 6, 75–95, 1993a.
Giorgi, F., M. R. Marinucci, and G. T. Bates, Development of a second generation regional climate model (regcm2)
i: Boundary layer and radiative transfer processes, Mon. Wea. Rev., 121, 2794–2813, 1993b.
Giorgi, F., M. R. Marinucci, G. T. Bates, and G. DeCanio, Development of a second generation regional climate
model (regcm2) ii: Convective processes and assimilation of lateral boundary conditions, Mon. Wea. Rev., 121,
2814–2832, 1993c.
Giorgi, F., X. Q. Bi, and Y. Qian, Radiative forcing and regional climatic effects of anthropogenic aerosols over
East Asia: A regional coupled climate-chemistry/aerosols model study, J. Geophys. Res., 107, 2002.
Giorgi, F., X. Q. Bi, and Y. Qian, Indirect vs. direct effects of anthropogenic sulfate on the climate of east asia as
simulated with a regional coupled climate-chemistry/aerosol model, Climatic Change, 58, 345–376, 2003a.
Giorgi, F., R. Francisco, and J. S. Pal, Effects of a subgrid-scale topography and land use scheme on the simulation
of surface climate and hydrology. part 1: Effects of temperature and water vapor disaggregation., Journal of
Hydrometeorology, 4, 317–333, 2003b.
Giorgi, F., J. S. Pal, X. Bi, L. Sloan, N. Elguindi, and F. Solmon, Introduction to the tac special issue: The regcnet
network., Theoretical and Applied Climatology, 86, 1–4, 2006.
Grell, G., Prognostic evaluation of assumptions used by cumulus parameterizations, Mon. Wea. Rev., 121, 764–787,
1993.
Grell, G. A., J. Dudhia, and D. R. Stauffer, Description of the fifth generation Penn State/NCAR Mesoscale Model
(MM5), Tech. Rep. TN-398+STR, NCAR, Boulder, Colorado, pp. 121, 1994.
Grenier, H., and C. S. Bretherton, A moist pbl parameterization for large-scale models and its application to
subtropical cloud-topped marine boundary layers, Monthly Weather Review, 129, 357–377, 2001.
Hack, J. J., B. A. Boville, B. P. Briegleb, J. T. Kiehl, P. J. Rasch, and D. L. Williamson, Description of the ncar
community climate model (ccm2), Tech. Rep. NCAR/TN-382+STR, National Center for Atmospheric Research,
1993.
Henderson-Sellers, B., Calculating the surface energy balance for lake and reservoir modeling: A review, Rev.
Geophys., 24(3), 625–649, 1986.
Holtslag, A. A. M., and B. A. Boville, Local versus nonlocal boundary-layer diffusion in a global climate model,
J. Climate, 6, 1993.
Holtslag, A. A. M., E. I. F. de Bruijn, and H.-L. Pan, A high resolution air mass transformation model for shortrange weather forecasting, Mon. Wea. Rev., 118, 1561–1575, 1990.
Hostetler, S. W., G. T. Bates, and F. Giorgi, Interactive nesting of a lake thermal model within a regional climate
model for climate change studies, Geophysical Research, 98, 5045–5057, 1993.
29
Hsie, E. Y., R. A. Anthes, and D. Keyser, Numerical simulation of frontogenisis in a moist atmosphere, J. Atmos.
Sci., 41, 2581–2594, 1984.
Kiehl, J. T., J. J. Hack, G. B. Bonan, B. A. Boville, B. P. Breigleb, D. Williamson, and P. Rasch, Description of
the ncar community climate model (ccm3), Tech. Rep. NCAR/TN-420+STR, National Center for Atmospheric
Research, 1996.
Kueppers, L., et al., Seasonal temperature response to land-use change in the western united states., Global and
Planetary Change, 60, 2008.
Laurent, B., B. Marticorena, G. Bergametti, J. Leon, and N. Mahowald, Modeling mineral dust emissions from the
sahara desert using new surface properties and soil database, Journal of Geophysical Research, 113, d14218,
2008.
Lawrence, P., and T. Chase, Representing a new MODIS consistent land surface in the Community Land Model
(CLM3.0), J. Geophys. Res., 112, g01023, 2007.
O’ Brien, T. e. a., A new turbulence parameterization in regcm allows long-term mesoscale simulation of marine
stratocumuls., Climate Dynamics, 2011.
Oleson, K. e. a., Technical description of the Community Land Model (CLM), Tech. Rep. Technical Note
NCAR/TN-461+STR, NCAR, 2004.
Oleson, K. W., et al., Improvements to the Community Land Model and their impact on the hydrological cycle,
Journal of Geophysical Research-Biogeosciences, 113(G1), 2008.
Pal, J. S., E. E. Small, and E. A. B. Eltahir, Simulation of regional-scale water and energy budgets: Representation
of subgrid cloud and precipitation processes within RegCM, J. Geophys. Res.-Atmospheres, 105(D24), 29,579–
29,594, 2000.
Pal, J. S., F. Giorgi, X. Bi, et al., The ICTP RegCM3 and RegCNET: Regional climate modeling for the developing
world, Bull. Amer. Meteor. Soc., 88, 1395–1409, 2007.
Patterson, J. C., and P. F. Hamblin, Thermal simulation of a lake with winter ice cover, Limn. Oceanography, 33,
323–338, 1988.
Slingo, J. M., A gcm parameterization for the shortwave radiative properties of water clouds, J. Atmos. Sci., 46,
1419–1427, 1989.
Small, E. E., and L. C. Sloan, Simulating the water balance of the aral sea with a coupled regional climate-lake
model, J. Geophys. Res., 104, 6583–6602, 1999.
Solmon, F., M. Mallet, N. Elguindi, F. Giorgi, A. Zakey, , and A. Konare, Dust aerosol impact on regional
precipitation over western africa, mechanisms and sensitivity to absorption properties, Geophysical Research
Letters, 35, l24705, 2008.
Steiner, A. L., J. S. Pal, S. A. Rauscher, J. L. Bell, N. S. Diffenbaugh, A. Boone, L. C. Sloan, and F. Giorgi,
Land surface coupling in regional climate simulations of the West African monsoon, Climate Dynamics, 33(6),
869–892, 2009.
Sundqvist, H., E. Berge, and J. E. Kristjansson, The effects of domain choice on summer precipitation simulation
and sensitivity in a regional climate model, J. Climate, 11, 2698–2712, 1989.
Tiedtke, M., A comprehensive mass flux scheme for cumulus parameterization on large scale models, Mon. Wea.
Rev., 117, 1779–1800, 1989.
Zeng, X., A prognostic scheme of sea surface skin temperature for modeling and data assimilation, Geophysical
Research Letters, 32, l14605, 2005.
Zeng, X., M. Zhao, and R. E. Dickinson, Intercomparison of bulk aerodynamic algoriths for the computation of
sea surface fluxes using toga coare and tao data, J. Climate, 11, 2628–2644, 1998.
30
BATS Biosphere-Atmosphere Transfer Scheme
BATS1e Biosphere-Atmosphere Transfer Scheme version 1e
CAM Community Atmosphere Model
CAPE convective available potential energy
CCM Community Climate Model
CCM1 Community Climate Model version 1
CCM2 Community Climate Model version 2
CCM3 Community Climate Model version 3
CLM Community Land Surface Model
CLM0 Common Land Model version 0
CLM2 Community Land Model version 2
CLM3 Community Land Model version 3
CMAP CPC Merged Analysis of Precipitation
CRU Climate Research Unit
CPC Climate Prediction Center
ECMWF European Centre for Medium-Range Weather Forecasts
ERA40 ECMWF 40-year Reanalysis
ESMF Earth System Modeling Framework
ESP Earth Systems Physics
FAO Food and Agriculture Organization of the United Nations
fvGCM NASA Data Assimilation Office atmospheric finite-volume general circulation model
GLCC Global Land Cover Characterization
GCM General Circulation Model
HadAM3H Hadley Centre Atmospheric Model version 3H
ICTP Abdus Salam International Centre for Theoretical Physics
IPCC Intergovernmental Panel on Climate Change
IBIS Integrated BIosphere Simulator
LAI leaf area index
LAMs limited area models
LBCs lateral boundary conditions
MC2 Mesoscale Compressible Community model
MIT Massachusetts Institute of Technology
MM4 Mesoscale Model version 4
MM5 Mesoscale Model version 5
31
MERCURE Modelling European Regional Climate Understanding and Reducing Errors
NNRP NCEP/NCAR Reanalysis Product
NNRP1 NCEP/NCAR Reanalysis Product version 1
NNRP2 NCEP/NCAR Reanalysis Product version 2
NCAR National Center for Atmospheric Research
NCEP National Centers for Environmental Prediction
PBL planetary boundary layer
PC Personal Computer
PIRCS Project to Intercompare Regional Climate Simulations
PFT plant functional type
PSU Pennsylvania State University
PWC Physics of Weather and Climate
RCM Regional Climate Model
RegCM REGional Climate Model
RegCM1 REGional Climate Model version 1
RegCM2 REGional Climate Model version 2
RegCM2.5 REGional Climate Model version 2.5
RegCM3 REGional Climate Model version 3
RegCM4 REGional Climate Model version 4
RegCNET REGional Climate Research NETwork
RMIP Regional Climate Model Intercomparison Project
ROMS Regional Oceanic Modeling System
SIMEX the Simple EXplicit moisture scheme
SST sea surface temperature
SUBEX the SUB-grid EXplicit moisture scheme
USGS United States Geological Survey
JJA June, July, and August
JJAS June, July, August, and September
JFM January, February, and March
32

advertisement

Key Features

  • Open Source
  • User Friendly
  • Portable Code
  • Wide range of applications
  • Community model
  • Publicly available

Frequently Answers and Questions

What are the main components of the RegCM modeling system?
The RegCM modeling system has four components: Terrain, ICBC, RegCM, and Postprocessor. Terrain and ICBC are the two components of RegCM preprocessor.
What is the vertical coordinate of the model?
The model's vertical coordinate is terrain-following, meaning that the lower grid levels follow the terrain while the upper surface is flatter.
What are the different map projections available in the model?
The model has a choice of four map projections: Lambert Conformal, Polar Stereographic, Normal Mercator, and Rotated Mercator.

Related manuals

Download PDF

advertisement