CLIMATE CLASSIFICATION

Climate classification, the formalization of systems that recognize, clarify, and simplify climatic
similarities and differences between geographic areas in order to enhance the scientific
understanding of climates. Such classification schemes rely on efforts that sort and group vast
amounts of environmental data to uncover patterns between interacting climatic processes. All
such classifications are limited since no two areas are subject to the same physical or biological
forces in exactly the same way. The creation of an individual climate scheme follows either a
genetic or an empirical [Link] climate of an area is the synthesis of the environmental
conditions (soils, vegetation, weather, etc.) that have prevailed there over a long period of
time. This synthesis involves both averages of the climatic elements and measurements of
variability (such as extreme values and probabilities). Climate is a complex, abstract concept
involving data on all aspects of Earth’s environment. As such, no two localities on Earth may be
said to have exactly the same climate.

Nevertheless, it is readily apparent that, over restricted areas of the planet, climates vary within
a limited range and that climatic regions are discernible within which some uniformity is
apparent in the patterns of climatic elements. Moreover, widely separated areas of the world
possess similar climates when the set of geographic relationships occurring in one area parallels
that of another. This symmetry and organization of the climatic environment suggests an
underlying worldwide regularity and order in the phenomena causing climate (such as patterns
of incoming solar radiation, vegetation, soils, winds, temperature, and air masses). Despite the
existence of such underlying patterns, the creation of an accurate and useful climate scheme is
a daunting [Link], climate is a multidimensional concept, and it is not an obvious decision as
to which of the many observed environmental variables should be selected as the basis of the
classification. This choice must be made on a number of grounds, both practical and theoretical.
For example, using too many different elements opens up the possibilities that the classification
will have too many categories to be readily interpreted and that many of the categories will not
correspond to real climates. Moreover, measurements of many of the elements of climate are
not available for large areas of the world or have been collected for only a short time. The
major exceptions are soil, vegetation, temperature, and precipitation data, which are more
extensively available and have been recorded for extended periods of time.

KOPPEN'S CLASSIFICATION
Köppen’s classification is based on a subdivision of terrestrial climates into five major types,
which are represented by the capital letters A, B, C, D, and E. Each of these climate types except
for B is defined by temperature criteria. Type B designates climates in which the controlling
factor on vegetation is dryness (rather than coldness). Aridity is not a matter of precipitation
alone but is defined by the relationship between the precipitation input to the soil in which the
plants grow and the evaporative losses. Since evaporation is difficult to evaluate and is not a
conventional measurement at meteorological stations, Köppen was forced to substitute a
formula that identifies aridity in terms of a temperature-precipitation index (that is,
evaporation is assumed to be controlled by temperature). Dry climates are divided into arid
(BW) and semiarid (BS) subtypes, and each may be differentiated further by adding a third
code, for warm (h) or cold (k).

As noted above, temperature defines the other four major climate types. These are subdivided,
with additional letters again used to designate the various subtypes. Type A climates, the
warmest, are differentiated on the basis of the seasonality of precipitation: Af (no dry season),
Am (short dry season), or Aw (winter dry season). Type E climates, the coldest, are
conventionally separated into tundra (ET) and snow/ice climates (EF). The midlatitude C and D
climates are given a second letter, f (no dry season) or w (winter dry) or s (summer dry), and a
third symbol—a, b, c, or d (the last subclass exists only for D climates)—indicating the warmth
of the summer or the coldness of the winter. Although Köppen’s classification did not consider
the uniqueness of highland climate regions, the highland climate category, or H climate, is
sometimes added to climate classification systems to account for elevations above 1,500
metres (about 4,900 feet). The table gives the specific criteria for the Köppen-Geiger-Pohl
system of 1953.
CLASSIFICATION OF MAJOR CLIMATIC TYPES
ACCORDING TO THE MODIFIED KÖPPEN-
GEIGER SCHEME

LETTER SYMBOL
1st 2nd 3rd criterion

1In the formulas above, r is average annual precipitation total (mm), and t is average annual
temperature (°C). All other temperatures are monthly means (°C), and all other precipitation
amounts are mean monthly totals (mm).

2Any climate that satisfies the criteria for designation as a B type is classified as such,
irrespective of its other characteristics.

3The summer half of the year is defined as the months April–September for the Northern
Hemisphere and October–March for the Southern Hemisphere.

4Most modern climate schemes consider the role of altitude. The highland zone has been taken
from G.T. Trewartha, An Introduction to Climate, 4th ed. (1968).

Data Sources: Adapted from Howard J. Critchfield, General Climatology, 4th ed. (1983), and
M.C. Peel, B.L. Finlayson, and T.A. McMahon, "Updated World Map of the Köppen-Geiger
Climate Classification," Hydrology and Earth System Sciences, 11:1633–44 (2007).

A temperature of coolest month 18 °C or higher

f prec

precipitation in driest month less than 60 mm but equal to or greater than 100 – (r/25)1
w precipitation in driest month less than 60 mm and less than 100 – (r/25)

B2 70% or more of annual precipitation falls in the summer half of the year and r
less than 20t + 280, or 70% or more of annual precipitation falls in the winter half of the year
and r less than 20t, or neither half of the year has 70% or more of annual precipitation and r
less than 20t + 1403

W r is less than one-half of the upper limit for classification as a B type (see above)

S r is less than the upper limit for classification as a B type but is more than one-
half of that amount

h t equal to or greater than 18 °C

k t less than 18 °C

C temperature of warmest month greater than or equal to 10 °C, and

temperature of coldest month less than 18 °C but greater than –3 °C

s precipitation in driest month of summer half of the year is less than 30 mm and
less than one-third of the wettest month of the winter half

w precipitation in driest month of the winter half of the year less than one-tenth
of the amount in the wettest month of the summer half

f precipitation more evenly distributed throughout year; criteria for neither s nor
w satisfied

a temperature of warmest month 22 °C or above

b temperature of each of four warmest months 10 °C or above but warmest

month less than 22 °C

c temperature of one to three months 10 °C or above but warmest month less

than 22 °C
D temperature of warmest month greater than or equal to 10 °C, and temperature
of coldest month –3 °C or lower

s same as for type C

w same as for type C

f same as for type C

a same as for type C

b same as for type C

c same as for type C

d temperature of coldest month less than –38 °C (d designation then used instead
of a, b, or c)

E temperature of warmest month less than 10 °C

T temperature of warmest month greater than 0 °C but less than 10 °C

F temperature of warmest month 0 °C or below

H4 temperature and precipitation characteristics highly dependent on traits of

adjacent zones and overall elevation—highland climates may occur at any latitude.

The Köppen classification has been criticized on many grounds. It has been argued that extreme
events, such as a periodic drought or an unusual cold spell, are just as significant in controlling
vegetation distributions as the mean conditions upon which Köppen’s scheme is based. It also
has been pointed out that factors other than those used in the classification, such as sunshine
and wind, are important to vegetation. Moreover, it has been contended that natural
vegetation can respond only slowly to environmental change, so that the vegetation zones
observable today are in part adjusted to past climates. Many critics have drawn attention to the
rather poor correspondence between the Köppen zones and the observed vegetation
distribution in many areas of the world. In spite of these and other limitations, the Köppen
system remains the most popular climatic classification in use today.

A major contribution to climate grouping was made by the American geographer-climatologist

C. Warren Thornthwaite in 1931 and 1948. He first used a vegetation-based approach that
made use of the derived concepts of temperature efficiency and precipitation effectiveness as a
means of specifying atmospheric effects on vegetation. His second classification retained these
concepts in the form of a moisture index and a thermal efficiency index but radically changed
the classification criteria and rejected the idea of using vegetation as the climatic integrator,
attempting instead to classify “rationally” on the basis of the numerical values of these indices.
His 1948 scheme is encountered in many climatology texts, but it has not gained as large a
following among a wide audience as the Köppen classification system has, perhaps because of
its complexity and the large number of climatic regions it defines.

While vegetation-based climate classifications could be regarded as having relevance to human

activity through what they may indicate about agricultural potential and natural environment,
they cannot give any sense of how human beings would feel within the various climate types.
Terjung’s 1966 scheme was an attempt to group climates on the basis of their effects on human
comfort. The classification makes use of four physiologically relevant parameters: temperature,
relative humidity, wind speed, and solar radiation. The first two are combined in a comfort
index to express atmospheric conditions in terms perceived as extremely hot, hot, oppressive,
warm, comfortable, cool, keen, cold, very cold, extremely cold, and ultra cold. Temperature,
wind speed, and solar radiation are combined in a wind effect index expressing the net effect of
wind chill (the cooling power of wind on exposed surfaces) and addition of heat to the human
body by solar radiation. These indices are combined for different seasons in different ways to
express how humans feel in various geographic areas on a yearly basis. Terjung visualized that
his classification would find applicability in medical geography, climatological education,
tourism, housing, and clothing and as a general analytical [Link] other specialized empirical
classifications have been devised. For example, there are those that differentiate between
types of desert and coastal climates, those that account for different rates of rock weathering
or soil formation, and those based on the identification of similar agricultural climates.

THORNTHWAITE CLIMATE CLASSIFICATION

Thornthwaite climate classification System for describing climates, devised in 1931 and revised
in 1948 by the American climatologist C. W. Thornthwaite, that divides climates into groups
according to the vegetation characteristic of them, the vegetation being determined by
precipitation effectiveness (P/E, where P is the total monthly precipitation, and E is the total
monthly evaporation). The sum of the monthly P/E values gives the P/E index, which is used to
define five humidity provinces, with associated vegetation. A P/E index of more than 127 (wet)
indicates rain forest; 64–127 (humid) indicates forest; 32–63 (subhumid) indicates grassland;
16–31 (semi-arid) indicates steppe; less than 16 (arid) indicates desert. In 1948 the system was
modified to incorporate a moisture index, which relates the water demand by plants to the
available precipitation, by means of an index of potential evapotran-spiration (PE), calculated
from measurements of air temperature and day length. In arid regions the moisture index is
negative because precipitation is less than the PE. The system also uses an index of thermal
efficiency, with accumulated monthly temperatures ranging from 0, giving a frost climate, to
more than 127, giving a tropical [Link] the Köppen climate classification is the most
common climate classification in use today, the 1948 Thornthwaite classification is frequently
cited as an improved climate classification system for its rational approach. However, the
Thornthwaite classification is infrequently used because it tends to be too complex for use in
everyday settings and world maps of the classification were never produced. This paper will
present global maps of all four components of the 1948 Thornthwaite climate classification—a
long-time wish of John "Russ" Mather, to whom this paper is dedicated. In addition, a revised
Thornthwaite-type climate classification is presented with the intent of providing a more
rational climate classification for everyday use in a classroom setting. This classification uses an
amended version of the Thornthwaite moisture index, not only to delineate climatic moisture
gradients but also to define a single seasonality index responsive to mean seasonal variation in
both thermal and moisture conditions. Replacing the two cumbersome seasonality indicators in
the original Thornthwaite classification with one variable greatly improves the utility of the
classification. Results from this classification are compared to the Köppen and original
Thornthwaite climate classification schemes.

FactorThornthwaite realized that precipitation (P) alone is not a good indicator of mois-ture
conditions in an environment. He considered this a significant drawback tomost climate
classification systems. Instead, he developed the concept of potentialevapotranspiration (PE),
derived from temperature and day length, to estimate thewater need of plants in a given
environment. Using PE in combination with P, hedeveloped his water budget methodology to
create a moisture index. Unlike othermoisture indices, however, Thornthwaite derived his index
from separately calcu-lated humidity (Ih) and aridity (Ia) indices, based on moisture surplus and
deficitcalculations from the water budgetIh = 100 S/PE andIa = 100 D/PE (1)where S is the water
surplus and D is the water [Link] suggested that perennial plants are sufficiently
deeply rooted to beable to access surplus moisture that percolates below the soil layer, thus
minimizingthe effects of drought. For this reason, he decided that 6 inches (~15 cm) of
watersurplus was sufficient to offset 10 inches (~25 cm) of moisture deficit for deep-
rootedvegetation and subsequently developed the following weighted moisture index:

444JOHANNES J. FEDDEMAIm = Ih – 0.6Ia(2)Values for this index range from -60 to infinity and
Thornthwaite used valuesfrom -60 to 100 in rational increments of 20 to classify climates into
humidityclasses, labeled in a similar fashion to the Köppen system with capital letters (Table1).
The Thornthwaite and Mather (1955) revision of the water budget methodologyimplied a
modification of this index so that the weighting of deficit and surplusbecame the same.
Removing the 0.6 weighting the moisture index was simplifiedto eitherIm = 100[(S – D)/PE]
orIm = 100(P/PE – 1) (3)With this formulation, the classification lost its rational increment of 20
since thethree dry climate types were now bound by -33, -66, and -100 (Thornthwaite
andMather, 1955). Figure 1 presents a global map of the moisture factor classificationusing the
original 1948 moisture [Link] the 1955 revision, Mather spent considerable time
focusing on twoquestions. First, he felt that the 1955 modification defeated the original
concept ofa “rational” classification (i.e., the use of even class intervals) and he
frequentlydebated the validity of the modified scheme in this context. Second, while the mois-
ture index has traditionally been applied on an annual time scale, he debated thevalidity of
using the index on a sub-annual time scale, for example, on monthly orseasonal time scales. As
will be shown, use of a sub-annual time scale allows themoisture index to be manipulated to
provide significant additional informationabout a [Link] Thermal FactorThornthwaite
based his thermal factor, called the Index of Thermal Efficiency, onPE. Most climate
classifications use temperature as an indicator of thermalefficiency. However, Thornthwaite
recognized that temperature alone was not nec-essarily an adequate indicator of the
productivity of an environment, especially forTable 1. Thornthwaite Moisture Derived Climate
TypesClimatic type Moisture indexA Perhumid 100 and aboveB4 Humid 80 to 100B3 Humid 60
to 80B2 Humid 40 to 60B1 Humid 20 to 40C2 Humid 0 to 20C1 Humid -20 to 0D Semiarid -40 to
-20E Arid -60 to -40.
 MILLER'S CLIMATE CLASSIFICATION
The Miller Classification is a technique that meteorologists use to classify nor'easters. The
system splits nor'easters into five categories: Miller A, Miller B, Miller C, Miller D, and Miller E;
the classification system initially started out with the first two categories. The system was
derived by meteorologist and researcher J.E. Miller in 1946.

The Miller Classification was created by meteorologist and researcher J.E. Miller in 1946.
Meteorologists use the technique to determine the track and severity of nor'easters. Storms
that receive the Type A classification develop primarily in the Gulf of Mexico or along the
southern East Coast, near Georgia and South Carolina. These storms develop primarily on the
Gulf Coast or East Coast along a decaying cold front, or along the marine/land airmass contrast
found on the East Coast. Miller A systems then quickly deepen and intensify as they move
northeastward. Type A storms typically move rapidly, hitting the Mid-Atlantic United States the
hardest. Nevertheless, New England can still receive significant snow depending on the
system's intensity. The Superstorm of 1993 is considered to have been a Miller Type-A storm.

Storms that receive the Type B classification develop inland over the United States. Storms that
come in from the west (up the Ohio Valley) are usually referred to as "Miller Type-B" storms.
These storms originate as an area of low pressure creating storming weather over the
Midwestern United States and the Ohio River Valley. These storms have a defined low-level
circulation center that moves toward the Appalachian Mountains from the west. As these
storms approach the mountains, they lose their defined circulation center due to high terrain.
[6] A more powerful cyclone spins up along the warm Gulf Stream waters off the coast of North
[Link] this re-development occurs, the storm produces precipitation, including heavy
snow, along the inland parts of the Mid-Atlantic. After redevelopment, the nor'easter takes a
northerly track, thenThe Miller Classification was created by meteorologist and researcher J.E.
Miller in 1946. Meteorologists use the technique to determine the track and severity of
nor'easters.

Storms that receive the Type A classification develop primarily in the Gulf of Mexico or along
the southern East Coast, near Georgia and South Carolina. These storms develop primarily on
the Gulf Coast or East Coast along a decaying cold front, or along the marine/land airmass
contrast found on the East Coast. Miller A systems then quickly deepen and intensify as they
move northeastward. Type A storms typically move rapidly, hitting the Mid-Atlantic United
States the hardest. Nevertheless, New England can still receive significant snow depending on
the system's intensity. The Superstorm of 1993 is considered to have been a Miller Type-A
storm.

Storms that receive the Type B classification develop inland over the United States. Storms that
come in from the west (up the Ohio Valley) are usually referred to as "Miller Type-B" storms.
These storms originate as an area of low pressure creating storming weather over the
Midwestern United States and the Ohio River Valley. These storms have a defined low-level
circulation center that moves toward the Appalachian Mountains from the west. As these
storms approach the mountains, they lose their defined circulation center due to high terrain. A
more powerful cyclone spins up along the warm Gulf Stream waters off the coast of North
Carolina. When this re-development occurs, the storm produces precipitation, including heavy
snow, along the inland parts of the Mid-Atlantic. After redevelopment, the nor'easter takes a
northerly track, then turns out to sea near New England. An example of this storm is the
February 5–6, 2010 North American blizzard.
Additional classification types

A study written by Albright and Cobb (2004) showed that there are five predominant patterns
that produce four inches or more of snowfall across the Mid-Atlantic. They added classification
types C through E, adding onto the Miller [Link] out to sea near New England. An
example of this storm is the February 5–6, 2010 North American blizzard.