Types of Models

1. Physical Models: Scale model replicas of observed ground surface characteristics

(e.g., topography, orography, buildings, etc) are constructed and inserted into a
chamber such as a wind tunnel (water tanks are also used). The flow of air or
other gases or liquids in this chamber is adjusted so as to best represent the
larger scale observed atmospheric conditions.
2. Mathematical Models: They utilizes basic analysis techniques as algebra and
calculus to solve directly all or a subset of 7 basic equations.

Mathematical models are further divided into two types

1. Diagnostic Models
2. Prognostic Models
(i) prognostic models utilize time‐dependent partial differential equations (which
describe the behavior of fluid or gaseous flow) to integrate forward in time.
(ii) diagnostic models either eliminate the time dependency completely from the
equations, or use the time derivative term over one finite time step. This type of
model does not integrate forward in time.
Diagnostic Models‐ Drawbacks

A major problem with this particular type of model, however, is that part of the wind flow
must be known before a simulation can be performed. Either observations can be made
and the model simulation forced to correspond closely to those measurements in their
vicinity or the disturbance to the wind due to the complex terrain is assumed to disappear
at some level.

Unfortunately, however, unless extensive data are available at the surface and
aloft, corresponding to the vertical and horizontal scales of the mesoscale system of
interest, such a representation will be necessarily inaccurate.

In the absence of meteorological observations, however, knowledge of terrain‐forcing can

still permit diagnostic model simulations, as long as all of the following are fulfilled:
(i) a strong wind flow exists in the lower troposphere;
(ii) a strong lid (i.e., strong inversion) occurs below the level of the highest terrain;
(iii) the boundary layer below the inversion is well mixed;
(iv) differential heating of the terrain surface is minimal.

Except for specific application, diagnostic models are limited in their applicability.
Prognostic models of airflow in complex terrain have been of two types:
‐those which simulate circulations generated by differential heating in complex terrain
(i.e., mountain‐valley flow, sea and land breezes)
‐those which simulate perturbations to the larger scale flow due to air advecting over
irregular terrain (i.e., forced airflow over rough terrain).

The dominant forcings of these two types of systems are distinct from one another.

For sea‐land breeze circulations (which are closely related to mountain‐valley flows),
the flow pattern becomes very nonlinear when the differential heating becomes
significant. For this case, heating at the surface is mixed upward by turbulence resulting
in a horizontal pressure gradient in the lower troposphere.
In response to this pressure gradient, accelerations result which can concentrate the
region of heating, resulting in an even stronger horizontal pressure gradient force.
Only as a result of increased frictional dissipation as these velocities increase, or when
the heating is removed, is this positive feedback stopped. Without extensive
observations so that the flow reversal associated with this mesoscale system can be
resolved, diagnostic models cannot represent this feature.
 Atmospheric Numerical Models
• Think of a mathematical model as a set of equations that are initialized and
then solved through time.
• The quality of the initialization is going to depend on the input (weather data)
and how realistic the equations are.
• Weather data/observation is imperfect.
 First, it is impossible to know weather information at every point. The
weather information is spread out by many miles and often hundreds of
miles between each other (i.e. weather balloon launch points). Thus,
there are huge gaps of weather information.
 Second, any sensor that is experiencing error, even very minor error, will
contaminate the data that goes into the forecast model.
 There are several additional reasons for imperfect weather data such as
unrepresentativeness error and the missing of mesoscale processes.

• All the models do not have the same data input; and have a different
mathematical way to solve the equations. There are also differences in
resolution (horizontal and vertical), and how physical processes are integrated
into the model.
Model Equations
Primitive Equations (PE) – the basic set of equations derived by using minimal
assumptions

Prognostic Equations:
 Horizontal momentum equations (u and v)
 Vertical momentum equation (w)
 Thermodynamic equation (θ)
 Moisture equation (q)

Diagnostic Equations
 Mass continuity equation
 Equation of state (ideal gas)
 Sometimes the vertical momentum equation is replaced by the
diagnostic hydrostatic equation in models of the large-scale flow.
 Coordinate systems
Horizontal coordinates: Horizontal position may be expressed directly in geographic
coordinates (latitude and longitude) for global models or in a map projection planar
coordinates for regional models.

Vertical coordinates:
The vertical coordinate is handled in various ways.
1. Geometric height (z) as the vertical coordinate.
2. Pressure coordinate system, in which the geopotential heights of constant‐pressure
surfaces become dependent variables, greatly simplifying the primitive equations.
This correlation between coordinate systems can be made since pressure decreases
with height through the Earth's atmosphere.
The first model used for operational forecasts, the single‐layer barotropic model, used a
single pressure coordinate at the 500‐millibar (about 5,500 m (18,000 ft)) level, and thus was
essentially two‐dimensional.
3. High‐resolution models—also called mesoscale models—such as the Weather
Research and Forecasting (WRF) model tend to use normalized pressure coordinates
referred to as sigma coordinates. This coordinate system receives its name from the
independent variable σ {sigma } used to scale atmospheric pressures with respect to
the pressure at the surface, and in some cases also with the pressure at the top of
the domain.
 Horizontal grids of a Numerical model

Numerical models of the atmosphere are based on the physical laws of

fluids. Atmosphere is a continuous fluid. However, we will discretize
the atmosphere into grids.

Basic framework =
Spatial grid on which the
equations of physics are
represented

Red lines = lat/lon grid

Grid cell = smallest scale

that can be resolved but
many important process
occurs on sub‐grid scales
Atmospheric Model Grids (Horizontal)
Problem near the poles
where longitudes
converge

Regional focus

MPAS

SPHERICAL CENTROIDAL VORONOI GRID (HEXAGON)

Rectangular/square grid or Lat‐Long grid:
It is the most commonly used grid in the NWP models.
It is simple in nature but sufferes from “the polar
problem” where the lines of equal longitudes, known as
meridians, converge to points at the poles.
Poles are unique points and may cause violation of global
conservation laws within the model.
To maintain computational stability near the poles, small
integration time‐steps could be used, but at great
expense.

A rotated grid can overcome the polar problem for limited area models.
• In the late 60’s and early 70’s, the application of quasi‐uniform grids was
proposed as a method to avoid the polar problem of the grid‐point models. For
example, the Global Forecasting System (GFS) model has roughly a square grid
near the equator, a more rectangular grid in the mid‐latitudes, and a triangular
grid near the poles, eventually converging to a point at the poles.
• The other problem with the latitude‐longitude grid is the need for special filters
to deal with the pole singularities. They also do not scale well on massively
parallel computers.
 Triangular grids
Triangular grids are not used as often in models as are
rectangular grids.
One form of quasi‐uniform grid whose base element is a
triangle, is the spherical geodesic grid.
Icosahedral grids, first introduced in the 1960s, give almost
homogeneous and quasi‐isotropic coverage of the sphere.
The main advantage of the geodesic grid is that all the grid cells are nearly the
same size. The uniform cell size allows for computational stability even with finite
volume schemes.
Hexagonal grids
Similar to triangular grids, hexagonal grids are also not often used. In this method,
variables are calculated at each grid intersection between different hexagons, in
addition to being calculated in the center of the hexagonal grid.
The advantages of the hexagonal grid are:
(i) It is computationally faster,
(ii) Removes the polar problem.
(iii) Permits larger explicit time steps.
(iv) Conservation of quantities in finite volume formulation.
Grid staggering
• After the choice of distribution of grid points (i.e., rectangular, triangular,
hexagonal, etc.), the next step is to arrange the prognostic variables on the
grid.
• When all the prognostic variables are defined at the same point in a grid, it is
called an unstaggered grid. On the other hand, when prognostic variables are
defined at more than one point in a grid, it is called a staggered grid.
• Characteristically, a staggered grid has the values of the wind components at
separate points than the thermodynamic variables within the grid cell.
Consequently, a model’s resolution is defined as the average distance
between adjacent grid points with the same variables.
Staggering is not only performed in the horizontal direction, but also in the
vertical direction, as well as in time, and any combinations thereof.

Horizontal staggering
There are a variety of methods in which models calculate temperature, winds,
surface pressure, geopotential, and other meteorological variables within
their grids.
Five different types of grids were introduced by Arakawa and Lamb in 1977.
Although staggered grids have higher equivalent resolution than unstaggered
grids, they are also more complex. Overall, the C grid is becoming more popular
in recent times with the E grid its closest competitor.

The resolution of models does

not depend solely on the size or
shape of the grid but also the
location in which the model
calculates various atmospheric
variables. As the number of grid
cells increases, the differences in
effective resolution between
different grid types eventually
goes to zero.
Horizontal staggering
Of these grids, A is an unstaggered grid where the variables are defined at the
same points, e.g., at the center or at the corners of the grid.
Since all variables are defined at all the grid points in the A grid, it is easy to
construct a higher‐order accurate scheme. The main disadvantage is that the
differences are computed over a distance of 2Δx, and the adjacent points are not
coupled for the pressure and convergence terms.

Grids B through E are all staggered grids where the variables are defined at
different points.

In grid B, evaluation of the two sets of variables are at different points, e.g., one
might evaluate the velocities at the center of a grid and masses at the grid
corners. Since the B and E grids have wind components at the same point, they
are often called semi‐staggered.
The B grid was chosen in the UK Met Office (UKMO) unified model [19, 20] for
climate simulation as well as numerical weather prediction.
In grid C, velocities are calculated at the mid‐point between
grid cells and h is calculated at the corners. The main
advantage of the C grid is that the pressure and convergence
terms are computed over a distance Δx, which is half of that in
the A grid indicating a doubling of the resolution compared to
the A grid. Most non‐hydrostatic mesoscale models like fifth
generation mesoscale model (MM5) and Weather Research
and Forecasting (WRF) use the C grid.

grid D is a slight variation of grid C, with the u and v variables being oriented with a
rotation of 90°. This variation allows for a simple evaluation of the geostrophic
wind. This is done by creating better averages for variables such as pressure
gradient, mass convergence/divergence, and the Coriolis terms. The D grid was
used (with time staggering) in NCEP’s nested grid model; however, this grid is no
longer used in any popular atmospheric model because of no added benefit.

The staggered E grid is rotated 45° relative to the B grid, but has an increased grid‐
spacing compared to the B grid. One problem with this grid is when the domain is
small and one‐dimensional, this grid is equivalent to grid A, but with less
computational efficiency. National Centers for Environmental Prediction (NCEP) eta
model uses grid E.
 Nested grids
• Some models are run with finer‐resolution grids nested inside coarser‐
resolution grids within the same model. Grid nesting is used when
computational limitations prohibit fine‐resolution grids from covering the entire
model domain.
• Nesting can be one‐way or two‐way. Information in a two‐way nested grid is
shared both ways, from the coarse‐grid to the fine grid and from the fine‐grid to
the coarse‐grid.
• The forecast variables for the coarse‐grid are updated based on the fine‐grid
prediction. The coarse‐grid prediction provides boundary conditions on the nest
interface for use in the fine‐grid prediction.

Nested grid with C grid staggering with

Example of a two way nested grid
3:1 grid size ratio
Nested grids
• Advantages of the two‐way nested grid include, fine‐scale processes
resolved on the finer grid are allowed to affect the larger‐scale flow on the
coarse grid. This is important for numerical weather prediction because the
small‐scale processes in the atmosphere greatly influence the large‐scale
processes in the atmosphere.
• Since the predictions on coarse‐resolution grids take less computer time and
memory compared to fine resolution grids, the outermost boundary of the
model can be moved far from the forecast region, while the fine‐resolution
domain remains small enough to run in real time.
• Moving nests are also common in the present models where a higher
resolution nest can move with the phenomenon of interest (e.g., hurricane)
to provide details that wouldn’t be possible in a coarse resolution
simulation.
 Vertical Coordinates and Resolution
• Properly depicting the vertical structure of the atmosphere leads to better
forecasts by Numerical Weather Prediction Models.
• To successfully understand this vertical structure, the model must have an
appropriate vertical coordinate with better resolution and thus better forecasts.
• Numerical Weather Prediction models produce forecasts by computing the
average over these coordinate surfaces, rather than on the surface itself.
• At this point, a forecaster might ask: "Why not use pressure and height surfaces,
as they are used in most maps anyway?" The reason that these surfaces are not
used in Numerical Weather Prediction is because they cause much confusion at
the ground. Therefore, other surfaces have been developed and used in the
vertical.
• Some of the most popular surfaces used in many of the current models are the
Sigma, Eta, and Theta surfaces.
Sigma Coordinate The sigma coordinate system defines the base at the model's ground level.
The surfaces in the sigma coordinate system follow the model terrain and are steeply sloped in
the regions where terrain itself is steeply sloped. the sigma coordinate system defines the
vertical position of a point in the atmosphere as a ratio of the pressure difference between
that point and the top of the domain to that of the pressure difference between a
fundamental base below the point and the top of the domain. Because it is pressure based
and normalized, it is easy to mathematically cast governing equations of the atmosphere into
a relatively simple form.
𝑃 𝑃 𝑃𝑡
Advantages 𝜎 𝑜𝑟 𝜎
𝑃𝑠 𝑃𝑠 𝑃𝑡

1) The sigma coordinate system conforms to natural terrain. This allows for good depiction
of continuous fields, such as temperature advection and winds, in areas where terrain
varies widely but smoothly.
2) It lends itself to increasing vertical resolution
near the ground. This allows the model to
better define boundary‐layer processes, such as
diurnal heating, low‐level winds, turbulence,
low‐level moisture, and static stability.

3) Eliminates the problem of vertical coordinate

systems intersecting the ground, unlike height
or isentropic coordinates.
Limitations
1) The model wind forecast depend upon accurate calculations of the pressure gradient
force (PGF). This is easly calculated in pressure coordinates when the height is known.
Yet, when sigma surfaces slope, the PGF must be expanded to include the effects of the
slope. This introduces errors because the lapse rate must be approximated at points that
lie in between the pressure surfaces where height is observed.
2) Sigma models have a difficult time dealing with weather events on the lee‐side of
mountain ranges (i.e. cold‐air damming, lee‐side cyclogenesis).
3) Because of the smoothing required in the mountain ranges along coastlines, land
points can be forced to extend beyond the true coastline.
Examples of Sigma Models or Varients
Aviation/Medium Range Forecast (AVN/MRF) Model
It has a vertical domain that runs from the surface to about 2.0 hPa. For a
surface pressure of 1000 hPa, the lowest level is at about 996 hPa.
The vertical domain is represented by a sigma coordinate.
The resolution is divided into 42 unequally spaced sigma levels, where for a
surface pressure of 1000 hPa, twelve levels are below 800 hPa, twenty levels
between 800hpa and 100hpa, and ten are above 100 hPa.
Nested Grid Model (NGM):
European Center Medium Range Weather Forecasting Model (ECMWF):
 Vertical Model Grid Model top:
WRF = 10 hPa

2.9 mbar
• Vertical resolution is also
important for quality of
Pure
simulations pressure
• Levels are not equally region

spaced (levels are closer 83 mbar

near surface and near
tropopause where rapid Hybrid
changes occurs) sigma‐pressure
region
• In WRF/HWRF: “hybrid”
coordinate ~ 985 mbar
‐ bottom: sigma coordinate (follows
topography) Pure sigma
region
‐ top: pressure coordinate Surface ~ 1000 mbar
‐ middle: hybrid sigma‐pressure
 σ coordinate system

The (x,y,p,t) coordinate system possesses certain advantages over the (x,y,z,t) system,
particularly, w.r.t the representation of the horizontal pressure force. However, the lower
boundary conditions offers some difficulties and is usually approximated over level terrain
by assuming ω 0 and over uneven terrain by ω 𝑣. 𝛻𝑝 where 𝑝 is the pressure at
bottom surface/ terrain level.

The 𝜎 coordinate system which is closely related to p‐system avoids these difficulties at the
lower boundary. Define a new vertical coordinate system 𝜎 as follows:
𝑝
𝜎
𝑝
Where p is pressure at any level and 𝑝 is surface pressure.
Y
At the surface 𝜎 1 𝑝 𝑝
𝑝 𝜙
At the top of the atmosphere 𝜎 0 𝑝 𝑝 0

𝜙 𝜙 𝜙 𝜙 𝜙 𝜙
From figure: 𝜙 𝜙 X
𝛿𝑥 𝛿𝑥 𝛿𝑥
𝜙 𝜙 𝜙 𝜙 𝜙 𝜙 𝛿𝜎 𝛿𝜎 𝛿𝜎 𝜕 𝑝
. 𝛿𝑥 𝛿𝑥 𝜕𝑥 𝑝
𝛿𝑥 𝛿𝑥 𝛿𝜎 𝛿𝑥
𝑝.
𝜙 𝜙 𝜙 𝜙 𝜕𝜙 𝛿𝜎
.
𝛿𝑥 𝛿𝑥 𝛿𝜎 𝛿𝑥

𝜙 𝜙 𝜙 𝜙 𝜕𝜙 𝜎 𝜕𝑝
.
𝛿𝑥 𝛿𝑥 𝜕𝜎 𝑝 𝜕𝑥

𝜕𝜙 𝜕𝜙 𝜕𝜙 𝜎 𝜕𝑝
.
𝜕𝑥 𝜕𝑥 𝜕𝜎 𝑝 𝜕𝑥

𝜕𝜙 𝜕𝜙 𝜕𝜙 𝜎 𝜕𝑝
.
𝜕𝑥 𝜕𝑥 𝜕𝜎 𝑝 𝜕𝑥

𝜎 𝜕𝜙
𝛻𝜙 𝛻𝜙 . 𝛻𝑝 .
𝑝 𝜕𝜎

𝜎 𝜕
𝛻 𝛻 . 𝛻𝑝 .
𝑝 𝜕𝜎
ETA Coordinate Model The fundamental base in the eta system is not at the ground surface, but
instead is at mean sea level. The eta coordinate system has surfaces that remain relatively horizontal at
all times. At the same time, it retains the mathematical advantages of the pressure based system that
does not intersect the ground. It does this by allowing the bottom atmospheric layer to be represented
within each grid box as a flat "step". The eta coordinate system defines the vertical position of a point
in the atmosphere as a ratio of the pressure difference between that point and the top of the domain
to that of the pressure difference between a fundamental base below the point and the top of the
domain. The ETA coordinate system varys from one at the base to zero at the top of the domain.
Because it is pressure based and normalized, it is easy to mathematically cast governing equations of
the atmosphere into a relatively simple form.

Advantages
1) Eta models do not need to perform the vertical interpolations that are necessary to calculate the
PGF in sigma models (Mesinger and Janji 1985). This reduces the error in PGF calculation and improves
the forecast of wind and temperature and moisture changes in areas of steeply sloping terrain.

2) Although the numerical formulation near the surface

is more complex, the low‐level convergence in areas of
steep terrain are far more representative of real
atmospheric conditions than in the simpler
formulations in sigma models (Black 1994). The
improved forecasts of low‐level convergence result in
better precipitation forecasts in these areas. The
improved predictable flow detail compared to a
comparable sigma model more than compensates for
the slightly increased computer run time.
 3) Compared with sigma models, eta models can often improve forecasts of cold air outbreaks,
damming events, and leeside cyclogenesis For example, in cold‐air damming events, the inversion
in the real atmosphere above the cold air mass on the east side of a mountain are preserved
almost exactly in an eta model.

Limitations
1) The step nature of the eta coordinate makes it difficult to retain detailed vertical structure in
the boundary layer over the entire model domain, particularly over elevated terrain.
2) Eta models do not accurately depict gradually sloping terrain. Since all terrain is represented
in discrete steps, gradual slopes that extend over large distances can be concentrated within as
few as one step. This unrealistic compression of the slope into a small area can be
compensated, in part, by increasing the vertical and/or horizontal resolution.
3) Eta models have difficulty predicting extreme downslope wind events.

An example of ETA Step Models

1. ETA Model:
This model uses 50 vertical levels (NCEP 2000).
This coordinate system makes the eta surfaces quasi‐horizontal everywhere as opposed to
sigma surfaces which can be steeply sloped(Black, 1994).
This model is often being updated, and changes are made quite frequently on its resolution.
 Vertical staggering
Staggering of grids in the vertical direction also provides certain advantages. For example,
vertical staggering introduced by Lorenz maintains the requirement of boundary conditions
of no flux at the top and bottom (Figure (a)). However, the Lorenz grid allows the formation
of a spurious computational mode.
This problem does not exist in the Charney‐Phillips grid; Figure (b)) in which the vertical
staggering, being more consistent (compared to Lorenz grid) with hydrostatic equation, does
not allow the additional computational mode. Most of the present state‐of‐the‐art
numerical models have staggered grids in the vertical direction with prognostic variables at
the center of the layer and the vertical velocity at the boundary of the layers.

Vertically staggered grid in the

WRF model
Conversion of Vertical co‐ordinates (Constant Surfaces):
Now we will consider other co‐ordinate systems which are more or less peculiar to
hydrodynamics and in particular to meteorology. The essential feature of these systems is
the replacement of the vertical co‐ordinate Z by pressure (p), or potential temperature (θ).
Now consider an arbitrary surface:
Let us consider an arbitrary property ϕ 𝑥, 𝑦, 𝑧, 𝑡 and a small section of the
surface ϕ = Constant at a time t = t0.
• Thus, a given geopotential gradient implies the same geostrophic wind at any height,
whereas a given horizontal pressure gradient implies different values of the geostrophic
wind depending on the density.

Pressure Coordinate Systems (or Isobaric coordinates)

• In the isobaric coordinate system, the horizontal pressure gradient force is
measured by the gradient of geopotential at constant pressure.
• Density no longer appears explicitly in the pressure gradient force; this is a
distinct advantage of the isobaric system. No density term in continuity
equation
Time staggering
Staggering of grids is not confined in space, staggering can also be in time. For
example, for atmospheric flow using the leapfrog scheme grid D is ideal when
staggered in time. Time staggering was first introduced by Eliassen, which
involves defining variables at every second time step on an offset D grid.
A slight variation of this approach performed by Bratseth used a higher‐order
interpolation to transfer values back from the offset grid to the original grid. All
the differences are calculated on a distance Δx. Despite this advantage, such time
staggering is not used due to the complexities that arise from the additional
staggering and need of special procedures for starting the leapfrog scheme.

Staggered grids in space and time with

(a) even time steps (b) odd time steps
PRIMITIVE EQUATION MODELS
PRIMITIVE EQUATION MODELS

Modern numerical forecast models are based on a formulation of the

dynamical equations, referred to as the primitive equations, which is
essentially the formulation proposed by Richardson. The primitive
equations differ from the complete momentum equations (shown here) in
that the vertical momentum equation is replaced by the hydrostatic
approximation, and the small terms in the horizontal momentum
equations are neglected.
In most models. σ‐coordinate system is used, and the vertical
dependence is represented by dividing the atmosphere into a number
of levels and utilizing finite difference expressions for vertical
derivatives.

Both finite differencing and the spectral method have been used for
horizontal discretization in operational primitive equation forecast
models.

Examples of such models are the grid point and spectral models
developed at the European Centre for Medium‐Range Weather
Forecasts (ECMWF), NCEP GFS, etc. The model is typically run at very
high resolution with the best estimate of the initial conditions.
 Hydrostatic Balance

Air is in hydrostatic equilibrium when the

upward‐directed pressure gradient force
is exactly balanced by the downward
force of gravity. Figure shows air in
hydrostatic equilibrium. Since there is no
net vertical force acting on the air, there is
Effective 
gravity
no net vertical acceleration, and the sum   
g  g r k   2 er  z s
of the forces is equal to zero
dw 1 p p
 -  2  u cos  - g   - g
dt  z z
The hydrostatic equation indicates the rate at which the pressure decreases with height is
equal to the air density times the acceleration of gravity. The minus sign indicates that, as
the air pressure decreases, the height increases.
Most of the time, the atmosphere is in hydrostatic balance, even when air slowly rises or
descends at a constant speed. However, this balance does not exist in violent
thunderstorms and tornadoes, where the air shows appreciable vertical acceleration.
 Scale Analysis (Vertical Motion)

For the vertical equation we can see that there are two terms
which are far greater than the other two.

The acceleration is of an order of magnitude so much smaller

than the Pgf and Gravity that it CAN be ignored

We can say therefore that for SYNOPTIC scale motion, vertical

acceleration can be ignored and that a state of balance called the
Hydrostatic Equation exists

dw 1 p
 -  2  u cos  -g
dt  z
10 -2 10 5
(10 -7
)  1 (101
)  10 -4
101
(10 -3
) - 101

10-5 10 4
 Geopotential and Geopotential height
Z
* Geopotential    g dz
0
is the specific work performed against gravity to displace a unit mass of air parcel
from the sea level to a height of Z.
 1 Z
* Geopotential height Z  
9 .8 9 .8  0
gdz
is nearly identical to geometric height, but has different physical meaning.
In some applications, the geometric height is converted to a quantity known as geopotential height. This is because “g” in the
equations and is the value at sea level. In reality, the value of g decreases with increasing distance from the center of the earth.
At the 10 km level, there is about a 0.05% difference between the value of g there and the value of g at sea level. The
calculations will result in an atmosphere that is slightly thicker than the use of the constant g would give. This difference is
insignificant for the discussions we are having in this class, but becomes more significant of an issue in other applications.

Hypsometric equation Rd  p1 
 z  z 2  z1  T v ln  
g  p2 
The Hypsometric Relation allows one to relate the thickness of a Rd Tv
slab of atmosphere bound by two isobars (say, the 1000 mb isobar H  ( scale height )
and the 500 mb isobar) to the mean virtual temperature of the layer. g
Virtual temperature Tv can be defined to include the effects of water vapor. It is the temperature at which dry air would have the
same density as the moist air, at a given pressure. In other words, two air samples with the same virtual temperature have the same
density, regardless of their actual temperature or relative humidity.
 The Hydrostatic Primitive Equations
• Simplified form of the equations of motion: the primitive equations

Momentum conservation:

Energy conservation:

Mass conservation:

Hydrostatic balance:
Source and sinks
Water vapor conservation due to phenomena occurring on
scales smaller than grid resolution
Parameterized processes
or “the physics”
 The hydrostatic primitive equations
• Simplified form of the equations of motion: the primitive equations

‐ Atmosphere is assumed to be in hydrostatic balance (good for horizontal

grid > 10 km)
compression due to gravity is balanced by a pressure gradient force
(involves ignoring acceleration in the vertical component of the
momentum equations)

‐ Earth is assumed to be spherical and some other small terms in the

momentum equations are neglected (atmosphere is thin compared to its
horizontal extent)
Scales of Atmospheric Processes

Important phenomena occurs at all scales.

Interactions between phenomena at
different scales
=> very challenging
 Physical parameterizations
 Process of including the effect of unresolved phenomena
 Usually based on:
Basic physics (law thermodynamics)
Empirical formulation from observations
 Key parameterizations in atmospheric model:
clouds
radiation
effects of unresolved turbulence and gravity waves
effects of convection on heat, moisture and momentum budgets.
 Behavior of model is critically dependent of these
parameterization processes
 Reduction of Pressure to Sea Level
In mountainous regions the difference in surface pressure from one measuring
station to another is largely due to differences in elevation.

To isolate that part of the pressure field that is due to the passage of weather systems, it is
necessary to reduce the pressures to a common reference level. For this purpose, sea
level is normally used.
Let Zg and pg be the geopotential and pressure at ground level and Z0 and p0 the
geopotential and pressure at sea level (Z0= 0).

Then, for the layer between the Earth’s surface and sea level, the hypsometric equation
becomes
 p0  Rd Tv
z g  z 0  z g  H ln   where H 
p  g0
 g
This can be solved to obtain the sea‐level pressure

 Zg   g0 Z g 
p0  p g exp    where p g exp  
H   Rd Tv 
The last expression shows how the sea-level pressure depends on the mean virtual
temperature between ground and sea level
 If Zg is small, the scale height H can be evaluated from the ground temperature.

 Zg  Zg
Also, if Zg ≤ H, the exponential can be approximated by exp    1 
H  H

Since H is about 8 km for the observed range of ground temperatures on Earth,

this approximation is satisfactory provided that Zg is less than a few hundred
meters.
 Zg  pg
With this approximation, we get p0  p g 1   or where p g  p0  Zg
 H  H

Since pg≈1000hPa and H≈8 km, the pressure correction (in hPa) is roughly equal
to Zg (in meters) divided by 8.

1
p g  p0  Z g
8

In other words, near sea level the pressure decreases by about 1 hPa for every 8
m of vertical ascent.