Skip to content

Köppen and Thornthwaite Classifications

CSIR NET Earth Sciences Part C: Köppen groups, letters and thresholds, Köppen types of India, Thornthwaite 1931 and 1948 schemes, worked classifications.

  • 19 min read
  • Updated
On this page · 12 sections
  1. Prerequisites
  2. The Hook: The Botanist’s Map
  3. Why Classify Climate?
  4. The Köppen Classification
  5. The Thornthwaite Classifications
  6. Memory Tricks
  7. Don’t Confuse
  8. Common Mistakes
  9. Practice
  10. Quick Revision Table
  11. Key Takeaways
  12. What’s Next

Prerequisites

The Hook: The Botanist’s Map

Why Classify Climate?

A classification compresses twelve months of temperature and precipitation into a symbol that says how a place behaves. Two approaches exist:

ApproachBasisExamples
Empirical (effective)observed effects, especially on vegetation; thresholds of temperature and rainfallKöppen, Thornthwaite, Trewartha
Geneticcauses: air masses, fronts, pressure beltsStrahler’s classification (low-, mid- and high-latitude climates controlled by air masses)

Köppen’s scheme was first published in full in 1900 and revised several times, with Rudolf Geiger, which is why it is often called the Köppen–Geiger classification. Thornthwaite published his first scheme in 1931 and his “rational” water-balance scheme in 1948.

The Köppen Classification

The five main groups

GroupNameDefining criterion (thermal or moisture)Vegetation it tracks
Atropical humidcoldest month mean 18 °C or abovemegatherm vegetation; no cold season
Bdryannual precipitation below a dryness threshold that rises with temperaturesteppe and desert
Cwarm temperate (mesothermal)coldest month between 0 °C (or −3 °C in Köppen’s original) and 18 °C; warmest month above 10 °Cmixed forests, evergreen broadleaf, Mediterranean scrub
Dsnow / cold (microthermal)coldest month 0 °C or below (−3 °C original); warmest above 10 °Cboreal forest (taiga)
Epolarwarmest month below 10 °Cno trees (10 °C warmest month approximates the treeline)

Many maps add H for highland climates, where altitude dominates. Group B is tested first: a hot place that is too dry is B, whatever its temperature.

The dryness threshold

With mean annual temperature \( T \) (°C) and annual precipitation \( P \) (mm), a widely used form (as in Peel and colleagues, 2007) is:

Seasonal rainfall regimeThreshold \( P_{th} \) (mm)
at least 70% of precipitation in the winter half-year\( 20T \)
at least 70% in the summer half-year\( 20T + 280 \)
neither\( 20T + 140 \)
$$\boxed{P < P_{th} \Rightarrow \text{B}; \qquad P < \tfrac{1}{2}P_{th} \Rightarrow \text{BW (desert)}, \text{ otherwise BS (steppe)}}$$

Summer rain raises the threshold because it falls when evaporation is highest and is less effective.

Second and third letters

LetterApplies toMeaning
fA, C, Dno dry season (in A: driest month at least 60 mm)
mAmonsoon: short dry season compensated by heavy rain; driest month below 60 mm but at least \( 100 - P/25 \) mm
wA, C, Ddry winter (in C and D: driest winter month below one-tenth of the wettest summer month)
sA, C, Ddry summer (in C and D: driest summer month below 40 mm and below one-third of the wettest winter month)
W / SBdesert / steppe
h / kBhot (mean annual temperature 18 °C or above) / cold (below 18 °C)
aC, Dhot summer: warmest month 22 °C or above
bC, Dwarm summer: warmest below 22 °C, at least four months at 10 °C or above
cC, Dcool summer: one to three months at 10 °C or above
dDvery cold winter: coldest month below −38 °C
T / FEtundra (warmest month 0 °C or above but below 10 °C) / ice cap (warmest month below 0 °C)
g(Indian usage)Ganges type: hottest month comes before the summer solstice and the rainy season
graph TD
    S["Station data"] --> B{"P below dryness threshold?"}
    B -->|yes| BW{"P below half the threshold?"}
    BW -->|yes| W["BW desert"]
    BW -->|no| BS["BS steppe"]
    B -->|no| E{"Warmest month below 10 C?"}
    E -->|yes| EE["E: ET or EF"]
    E -->|no| A{"Coldest month 18 C or above?"}
    A -->|yes| AA["A: Af, Am, Aw or As"]
    A -->|no| C{"Coldest month above 0 C?"}
    C -->|yes| CC["C: then f/s/w and a/b/c"]
    C -->|no| DD["D: then f/s/w and a/b/c/d"]

World distribution of the main types

TypeNameTypical regionsControlling belt
Aftropical rainforestAmazon, Congo basin, IndonesiaITCZ all year
Amtropical monsoonwest coast of India, Myanmar coastmonsoon
Awtropical savannamuch of peninsular India, East Africa, Brazilian camposITCZ in summer only
BWhhot desertSahara, Arabia, Thar, central Australiasubtropical high
BWk / BSkcold desert / steppeGobi, Central Asia, Ladakh, interior plateauscontinentality, rain shadow
Csa / CsbMediterraneanMediterranean basin, California, central Chile, southwestern Australia, the Capehigh in summer, westerlies in winter
Cfahumid subtropicalsoutheastern United States, eastern Chinaeast-coast warm currents, summer monsoon-like flow
Cfbmarine west coastnorthwest Europe, New Zealandwesterlies all year
Cwamonsoon-influenced warm temperatenorth India, southern Chinamonsoon
Dfb / Dfchumid continental / subarcticeastern Europe, southern Canada / Siberian and Canadian taigacontinentality
Dwdextreme subarcticnortheastern SiberiaSiberian High
ET / EFtundra / ice capArctic coasts / Greenland interior, Antarcticapolar high

Note that D climates are essentially confined to the Northern Hemisphere: the Southern Hemisphere has no large land mass at 40–70°.

Köppen climate types of India

The standard Köppen map of India (as given in NCERT’s India: Physical Environment) recognises:

TypeDescriptionRegion
Amwmonsoon with a short dry seasonwest coast south of Goa
Asmonsoon with a dry summerCoromandel coast of Tamil Nadu (rain mostly October–December)
Awtropical savannamost of the peninsular plateau south of the Tropic of Cancer
BShwsemi-arid steppenorthwest Gujarat, parts of western Rajasthan and Punjab
BWhwhot desertextreme western Rajasthan
Cwgmonsoon with a dry winter (Ganges type)Ganga plain, eastern Rajasthan, northern Madhya Pradesh, most of the northeast
Dfccold humid winter with a short summerArunachal Pradesh
Epolar (tundra and ice)high Himalaya of Jammu and Kashmir, Ladakh, Himachal Pradesh and Uttarakhand

Strengths and weaknesses

StrengthsWeaknesses
uses only monthly temperature and precipitation, available almost everywherethresholds are empirical, chosen from vegetation, not physics
boundaries match vegetation zones wellignores evaporation directly (only through \( T \) in the B threshold)
simple, global, easy to map from gridded datasays nothing about causes (air masses, winds)
widely used to map climate shifts under warmingsharp boundaries for gradual transitions; does not capture rainfall variability

Trewartha (1966, revised 1980) modified Köppen, redefining group C by the number of months at 10 °C or above, to make middle-latitude boundaries fit vegetation better.

The Thornthwaite Classifications

The 1931 scheme: precipitation and temperature efficiency

Thornthwaite’s first scheme used two indices computed from monthly precipitation \( P \) (inches) and temperature \( T \) (°F).

Precipitation-effectiveness (P-E) index:

$$\boxed{\text{P-E index} = \sum_{12} 115\left(\frac{P}{T - 10}\right)^{10/9}}$$

Temperature-efficiency (T-E) index:

$$\boxed{\text{T-E index} = \sum_{12} \frac{T - 32}{4}}$$
Humidity provinceP-E indexVegetation
A wet128 and aboverainforest
B humid64–127forest
C subhumid32–63grassland
D semi-arid16–31steppe
E aridbelow 16desert
Temperature provinceT-E index
A’ tropical128 and above
B’ mesothermal64–127
C’ microthermal32–63
D’ taiga16–31
E’ tundra1–15
F’ frost0

A third letter described the seasonal distribution of rainfall: r (rainfall adequate in all seasons), s (deficient in summer), w (deficient in winter), d (deficient in all seasons). The scheme still keyed classes to vegetation, which Thornthwaite himself later regarded as a weakness.

The 1948 “rational” scheme: the water balance

In 1948 Thornthwaite based the classification on potential evapotranspiration (PE), computed from temperature and day length (the formula is on the water-balance page), and on the monthly water balance giving the annual water surplus \( S \) and deficit \( D \).

IndexFormula
Humidity index\( I_h = 100\,S/PE \)
Aridity index\( I_a = 100\,D/PE \)
Moisture index\( I_m = I_h - 0.6\,I_a \) (1948); \( I_m = I_h - I_a = 100(S - D)/PE \) (Thornthwaite and Mather, 1955)

The factor 0.6 weights the deficit less heavily than the surplus in the 1948 index; the 1955 revision gave the two equal weight.

Moisture regions (1955 boundaries):

SymbolMoisture region\( I_m \)
Aperhumid100 and above
B4humid80 to 100
B3humid60 to 80
B2humid40 to 60
B1humid20 to 40
C2moist subhumid0 to 20
C1dry subhumid−33.3 to 0
Dsemi-arid−66.7 to −33.3
Earid−100 to −66.7

With the 1948 index the dry boundaries are −20 (C1/D), −40 (D/E) and −60.

Thermal efficiency is measured by annual PE itself, in cm:

SymbolThermal provinceAnnual PE (cm)
A'megathermal114 and above
B'4 to B'1mesothermal57 to 114 (four subdivisions of about 14.25 cm)
C'2, C'1microthermal28.5 to 57
D'tundra14.2 to 28.5
E'frostbelow 14.2

Two further letters complete the symbol: one for the seasonal variation of moisture (r, s, w, s2, w2 for moist climates, by the size and season of the deficit; d, s, w, s2, w2 for dry climates, by the size and season of the surplus), and one for the summer concentration of thermal efficiency (the share of annual PE falling in the three summer months: below 48% is a’, rising toward d’ above 88% in polar climates).

Köppen and Thornthwaite compared

KöppenThornthwaite (1948)
Basisvegetation boundaries matched to T and P thresholdswater balance: P compared with PE
Moisturedryness threshold from \( T \) and rainfall seasonmoisture index from surplus and deficit
Thermalmonthly temperature limits (18, 10, 0 or −3, −38 °C)annual PE
Data neededmonthly T and Pmonthly T, P, latitude (day length), soil storage
Best forworld maps, vegetation, teachingagriculture, irrigation, hydrology, drought
Weaknessempirical thresholds, no evaporationPE from temperature only (poor in hot, dry, windy climates); more computation

Memory Tricks

Don’t Confuse

PairDistinction
Köppen w vs sdry winter vs dry summer
Köppen m vs w (group A)short dry season offset by heavy rain vs pronounced dry season; the test is \( 100 - P/25 \)
BW vs BSbelow half the dryness threshold vs between half and the full threshold
h vs k (group B)mean annual temperature 18 °C or above vs below
C vs D boundarycoldest month 0 °C (modern) vs −3 °C (Köppen’s original)
Thornthwaite A (moisture) vs A’ (thermal)perhumid, \( I_m \ge 100 \) vs megathermal, \( PE \ge 114 \) cm
1931 vs 1948 ThornthwaiteP-E and T-E indices from P and T vs PE and the water balance
\( I_m \) 1948 vs 1955\( I_h - 0.6I_a \) vs \( I_h - I_a \)
Empirical vs genetic classificationeffects (vegetation, water balance) vs causes (air masses, fronts)

Common Mistakes

Practice

All questions below are original practice questions written for this page. They are not from any CSIR NET paper.

Level 1: Foundation

  1. What temperature criterion defines Köppen’s group E?
  2. What is the Köppen type of extreme western Rajasthan on the standard map of India?
  3. What quantity did Thornthwaite use in 1948 to measure thermal efficiency?
Answers
  1. Warmest month mean below 10 °C.
  2. BWhw, hot desert.
  3. Annual potential evapotranspiration (cm).

Level 2: Part B Style

Practice Q1. A coastal station has all months above 25 °C, annual rainfall 3,000 mm, and a driest month of 1 mm. Its Köppen type is:

(a) Af (b) Am (c) Aw (d) BSh

Answer: (b) Am

Not B: threshold \( 20\times 27 + 280 \approx 820 \) mm (taking a mean near 27 °C), far below 3,000. Coldest month above 18 °C: A. Driest month 1 mm < 60, so not Af. Am limit: \( 100 - 3000/25 = -20 \) mm; 1 mm ≥ −20, so Am. Heavy monsoon rain compensates for the dry season, the situation of India’s west coast.

Practice Q2. A station has a mean annual temperature of 26 °C and 600 mm of rain, 85% of it in the summer half-year. Its Köppen type is:

(a) BWh (b) BSh (c) Aw (d) BSk

Answer: (b) BSh

Summer rain: \( P_{th} = 20\times 26 + 280 = 800 \) mm. \( 600 < 800 \): group B. Half the threshold is 400; \( 600 > 400 \): BS. Mean temperature 26 °C ≥ 18: h. BSh. Using \( 20T + 140 = 660 \) would also give BS, but with \( 20T = 520 \) mm the rainfall exceeds the threshold and the station would wrongly fall outside group B.

Practice Q3. Match the Köppen type with its region in India.

TypeRegion
P. As1. Ganga plain
Q. Cwg2. Arunachal Pradesh
R. Dfc3. Coromandel coast
S. Amw4. West coast south of Goa

(a) P-3, Q-1, R-2, S-4 (b) P-4, Q-1, R-2, S-3 (c) P-3, Q-2, R-1, S-4 (d) P-1, Q-3, R-2, S-4

Answer: (a)

As: Tamil Nadu coast (dry high-Sun season). Cwg: Ganga plain. Dfc: Arunachal Pradesh. Amw: Konkan–Malabar coast.

Practice Q4. A station has \( PE = 900 \) mm, surplus 500 mm and deficit 50 mm. Using \( I_m = 100(S - D)/PE \), its moisture region is:

(a) B4 humid (b) B2 humid (c) C2 moist subhumid (d) A perhumid

Answer: (b) B2 humid

\( I_m = 100\times(500 - 50)/900 = 50 \), which lies in 40–60: B2. Using the surplus alone (\( I_h = 55.6 \)) would still give B2 here, but ignoring the deficit is a common slip that changes the class in drier stations.

Level 3: Part C Style

Practice Q5. Monthly data for three stations (all far from mountains):

StationColdest month (°C)Warmest month (°C)Mean annual T (°C)Annual P (mm)Rain seasonDriest month (mm)Wettest month (mm)
X1433251,00090% in summer5 (winter)300 (summer)
Y8241660075% in winter5 (summer)110 (winter)
Z−3215−835060% in summer10 (winter)70 (summer)

Which statements are correct?

  1. X is Cwa.
  2. Y is Csa.
  3. Z, which has two months at or above 10 °C, is Dfc, not E.
  4. Y would be BSk if its rain fell mainly in summer.

(a) 1 and 2 only (b) 1, 2 and 3 only (c) 1, 2, 3 and 4 (d) 2, 3 and 4 only

Answer: (b) 1, 2 and 3 only
  • X: threshold \( 20\times 25 + 280 = 780 < 1000 \): not B. Coldest 14 °C: between 0 and 18, warmest above 10: C. Driest winter month 5 < 300/10 = 30: w. Warmest 33 ≥ 22: a. Cwa. Statement 1 true.
  • Y: winter rain (75% ≥ 70%): \( P_{th} = 20\times 16 = 320 < 600 \): not B. Coldest 8 °C: C. Driest summer month 5 < 40 and < 110/3: s. Warmest 24 ≥ 22: a. Csa. Statement 2 true.
  • Z: neither half has 70%: \( P_{th} = 20\times(-8) + 140 = -20 \); 350 mm exceeds it, so not B. Warmest 15 °C > 10: not E. Coldest −32 ≤ 0 but above −38: D. Two months at or above 10 °C: c. Driest winter month 10 mm is not below one-tenth of the wettest summer month (7 mm), so not w; it the data show no dry summer (the driest month falls in winter): f. Dfc. Statement 3 true.
  • Statement 4 false: with summer rain \( P_{th} = 20\times 16 + 280 = 600 \); \( P = 600 \) is not below 600, so Y would still not be B (it would sit exactly on the boundary).

Practice Q6. In Thornthwaite’s 1931 scheme a station has six months with \( P = 4 \) in and \( T = 75 \) °F and six months with \( P = 0.5 \) in and \( T = 60 \) °F. Which classification is correct?

(a) P-E index about 35 (C, subhumid); T-E index about 107 (B’, mesothermal) (b) P-E index about 65 (B, humid); T-E index about 107 (B’) (c) P-E index about 35 (C); T-E index about 135 (A’, tropical) (d) P-E index about 18 (D, semi-arid); T-E index about 64 (B')

Answer: (a)

Wet months: \( 115\times(4/65)^{10/9} = 115\times 0.0451 = 5.19 \); six of them: 31.1. Dry months: \( 115\times(0.5/50)^{10/9} = 115\times 0.0060 = 0.69 \); six: 4.1. P-E index ≈ 35.3: C, subhumid (32–63).

T-E: wet months \( (75 - 32)/4 = 10.75 \), ×6 = 64.5; dry months \( (60 - 32)/4 = 7 \), ×6 = 42. T-E ≈ 106.5: B’, mesothermal (64–127).

Note how the dry months contribute almost nothing to the P-E index: the exponent and the temperature in the denominator penalise small rains.

Practice Q7. Two stations have the following annual water-balance values (mm):

StationPEAESD
M1,60070050900
N60059070010

Which statements are correct?

  1. M has \( I_m(1955) \approx -53 \): semi-arid (D), and is megathermal (A’).
  2. N has \( I_m(1955) \approx 115 \): perhumid (A).
  3. N’s annual precipitation is about 1,290 mm.
  4. Using the 1948 index, M would be classed arid (E).

(a) 1, 2 and 3 only (b) 1 and 2 only (c) 2, 3 and 4 only (d) 1, 2, 3 and 4

Answer: (a) 1, 2 and 3 only
  • 1 true: \( 100\times(50 - 900)/1600 = -53.1 \): between −66.7 and −33.3, D. PE 160 cm ≥ 114: A’. Check: \( AE + D = 1600 = PE \).
  • 2 true: \( 100\times(700 - 10)/600 = 115 \): A. (\( AE + D = 600 = PE \).)
  • 3 true: \( P = AE + S = 590 + 700 = 1{,}290 \) mm.
  • 4 false: \( I_h = 3.1 \), \( I_a = 56.3 \); \( I_m(1948) = 3.1 - 33.8 = -30.6 \): between −40 and −20, which is D (semi-arid), not E.

Quick Revision Table

IdeaFact or formula
Köppen groupsA (coldest ≥ 18 °C), B (dry), C (coldest 0/−3 to 18 °C), D (coldest ≤ 0/−3 °C), E (warmest < 10 °C)
Dryness threshold\( 20T \) (winter rain), \( 20T + 140 \), \( 20T + 280 \) (summer rain), mm
BW / BS\( P < P_{th}/2 \) / \( P_{th}/2 \le P < P_{th} \)
A subtypesAf driest ≥ 60 mm; Am driest ≥ \( 100 - P/25 \); Aw otherwise
s / wdriest summer < 40 mm and < 1/3 wettest winter / driest winter < 1/10 wettest summer
a / b / c / dwarmest ≥ 22; ≥ 4 months ≥ 10; 1–3 months ≥ 10; coldest < −38 °C
ET / EFwarmest 0–10 °C / below 0 °C
IndiaAmw west coast, As Coromandel, Aw plateau, BShw, BWhw Thar, Cwg Ganga plain, Dfc Arunachal, E Himalaya
Thornthwaite 1931P-E \( = \sum 115[P/(T-10)]^{10/9} \); T-E \( = \sum (T-32)/4 \); classes 128, 64, 32, 16
Thornthwaite 1948PE-based; \( I_m = I_h - 0.6I_a \) (1948), \( 100(S-D)/PE \) (1955)
Moisture regions (1955)A ≥ 100; B4–B1 80/60/40/20; C2 0–20; C1 −33.3–0; D −66.7 to −33.3; E below
Thermal (PE, cm)A’ ≥ 114; B’ 57–114; C’ 28.5–57; D’ 14.2–28.5; E’ < 14.2
Other schemesTrewartha (modified Köppen), Strahler (genetic, air masses)

Key Takeaways

What’s Next

Advertisement
Advertisement