Climate Classification

Surface data collected as described in Section 15.1 allow us to summarise the climate of a place in terms of average conditions and deviations from them. Particular attention is usually paid to monthly mean values of the daily-mean temperatures and daily extremes, along with the monthly rainfall.

The next step in condensing the flood of numbers that arise in continuously measuring the weather everywhere is to group homoclimes, places with similar climates. This imposes orderliness on the information, and stimulates wondering about the reasons for differences. Climate classification is also directly useful, e.g. in choosing crops for a particular climate—one selects from those grown in homoclimes. Or, to take a converse instance, there was the problem in Colombia of finding areas suited to the growing of tea, which became the problem of finding topoclimates in Colombia which resemble (in terms of monthly rainfall and mean temperature) those of tea-growing regions elsewhere in the world.

There are several ways of grouping similar climates, according to how we define 'similar'. Which criteria to use depends on the problem. If you are concerned with human comfort (Section 6.5), you would group places in terms of temperature and humidity. But rainfall would be more relevant if you are a farmer in an arid land. We will consider some of the criteria that have been used.

Latitude as the Criterion of Climatic Similarity

This is the oldest system, dating from the origin of the word 'climate', which is related to 'incline', as in the 'declination' of the Sun (Section 2.2). A tropical climate is found where the noonday Sun is inclined high in the sky.

Figures 3.4 and 10.3 indicate the effects of latitude on temperature and rainfall, by and large. But the effects of altitude and proximity to the sea (Sections 3.2 and 10.3) are ignored in using latitude alone as the criterion for classification.

Temperature as Criterion

Grouping places according to temperature was done in effect by annual-mean isotherms on maps of the world by Alexander von Humboldt in 1817, Alexander Supan in 1879 and Wladimir Koppen in 1884. An alternative is to categorise by the number of months when temperatures exceed the lower limit for plant growth, or the number of frost-free months, or the number of growing-degree days (Note 3.I). But this overlooks the importance of rainfall to plants.

Rain as Criterion

Alexsander Voeikov classified climates in 1874 according to the seasonal incidence of rain, leading to a world map in 1884, identifying areas of various degrees of wetness. Such a classification is useful in warm dry countries, where moisture is the main factor limiting growth. (Thus the preoccupation with rainfall in Australia, whereas Europeans are concerned about temperature and radiation.) An example is the Goyder Line in South Australia, which demarcated parts suited to growing wheat (Section 16.6). Elsewhere, the ratio of months with rainfall less than 60 mm to the number with over 100 mm has been used as a criterion for deciding where rice can be grown without irrigation.

An analagous criterion is aridity, a vague concept which takes in evaporation as well as rainfall. It may be expressed by the ratio of rainfall to the potential evaporation, P/E0. A climate is 'arid' (i.e. the soil is dry) when the ratio is low. For instance, a comparison of Figure 4.12 and Figure 10.12 indicates that the ratio is less than 0.1 in much of Australia's interior.

The temperature is often taken as proxy for the evaporation rate in assessing aridity, since T and Eo are closely related (Section 4.3). So P/Eo is replaced by the ratio of rainfall to temperature

P/T, where T is in degrees Celsius. But there are several other variants of either P/Eo or Eo/P, using net radiation or saturation deficit as a proxy for Eo. Categorisation of climates more explicitly in terms of soil moisture was introduced by Warren Thornthwaite in 1948, though it involved an unsatisfactory method of estimating the evaporation.

Vegetation as Criterion

Natural vegetation reflects the climate as a whole, and has been used as an indicator of it. Of course, vegetation also depends partly on the soil type, but that too is influenced by the climate. So coconut palms grow only where the monthly mean temperatures always exceed 18°C, pasture exists only where annual mean temperatures are above -2°C and trees where it is over +2°C, and the boundary of 'saltbush country' in Australia coincides with the isohyet of 200 mm during April to November. The association between vegetation and climate has led to the diagram in Figure 16.1. Temperature decreases vertically (a measure of increasing latitude or altitude) and rainfall increases diagonally, towards the lower right. As a result, TIP (or E/P, i.e. soil dryness) increases along the other diagonal, towards the lower left. Thus 'desert scrub' is found where rainfalls are between 125250 mm/a and the E/P ratio is above unity, i.e. the soil is dry on average.

The disadvantage of classifying climates according to the vegetation is that plants depend on other things also (competition, soil type, nutrient level, slope and so on), not only on climate. As a result, the classification is only approximate. Table 16.2 shows what a wide range of climates is associated with each kind of vegetation found in Australia. Thus, a climate with January/July rainfalls of 100/10 mm and temperatures 30/20°C can sustain either savanna woodland or tropical rainforest.

latitude polar

frigid

^/dry tundra \rnoist tundra)/ wet tundra V jjjp^j \ QOO

-0

alpine

■9 /

cold

£ J

desert \/ desert \J moist \V wet V/ pluvial A scrub A woodland A woodland A woodland

2000

subalpine

cool temperate

desert

\ / desert \/9rassland\/ \/ \/ X L, X or X moist forest X wet forest Y rainforest .4000 A scrub A dryscrub A A A /

montane

warm temperate low subtropical

16/ desert V

desert \J thorn \J. . . \J . \/ scrub X scrub X*Vforest YmoistforestYwetforest

,8000

4000 6

3000 12 17

1000 24

tropical scrub X woodland

/ arid forest V dry forest ymoist forest y wet forest y rainforest tropical tropical altitude: m temp: °C

4750 1.5

4500 3

4000 6

3000 12 17

1000 24

scrub X woodland

/ arid forest V dry forest ymoist forest y wet forest y rainforest tropical

Figure 16.1 Stratification of vegetation with altitude as a function of annual rainfall P. The ratio Eo/P is a measure of aridity.

Table 16.2 January and July conditions associated with various kinds of vegetation found in Australia, in descending order of January rainfall. The range shown in each box results from about ten values, whose rounded-off median is shown bold

Rainfall (mm/month) Monthly mean temperature (°Cj

Table 16.2 January and July conditions associated with various kinds of vegetation found in Australia, in descending order of January rainfall. The range shown in each box results from about ten values, whose rounded-off median is shown bold

Rainfall (mm/month) Monthly mean temperature (°Cj

Natural vegetation

January

July

January

July

Tropical rainforest

45-

-529

290

6-

128

35

21.7

-32.3

27

10.2-

-22.5

18

Savanna woodland

81-

-430

220

1-

39

8

22.4

-30.8

28

9.0-

25.5

23

Shrub savanna

63-

265

130

0-

36

3

26.9

-32.2

31

11.8

-22.2

19

Mallee

63

265

130

23

-104

55

18.5

-28.6

24

9.5-

14.3

11

Temperate rainforest

47-

147

90

84

-320

200

12.1

-15.4

13

2.1 -

9.4

4

Dry sclerophyll forest

10-

166

75

42

-174

70

16.7

-24.3

22

5.1 -

13.0

9

Savanna grasslands

45-

134

70

2-

22

10

29.6

-32.3

31

15.1

-22.0

17

Wet sclerophyll forest

17-

116

60

38

-185

60

14.9

-21.3

20

6.3-

13.7

10

Temperate woodland

11-

69

50

28

-93

45

14.5-

-26.8

23

4.8-

10.2

9

Alpine

37-

133

50

48

-258

100

13.3

-16.9

15

-2 to +8.4

6

Desert

20-

89

35

6-

46

15

26.7

-32.3

30

10.9

-18.1

13

Temperate semi-arid

10-

78

25

13

-25

15

25.5

-30.3

28

10.2

-13.0

11

steppe

Shrub steppe

11 -

28

20

9-

34

15

23.8

-31.3

25

8.7-

13.2

12

Koppen's Classification

Wladimir Koppen (1846-1940; his name is pronounced 'kerpen') first classified climates at the age of 24 and continued to refine his system until he was 90. His 1918 version is probably the most significant, and the most widely used nowadays. It involves grouping climates according to the kinds of vegetation present, using conditions at the boundaries of trees, for instance, as limits for the various categories (Note

16.A). As an example, South America's climates are shown in Figure 16.2.

About 20 per cent of the world's land has climates within class A (including 9 per cent as rainforest in 1987), 26 per cent in B, 15 per cent in C, 21 per cent in D and 17 per cent in E. D climates do not occur in the southern hemisphere; they characterise the climate on the large land masses between 45-65°N. Seventy-seven per cent of Australia's population live where the climate is labelled Cf, affecting 10

Isohyeten Kerpen
Figure 16.2 Koppen classes of climates of South America.

per cent of the continent's area, and another 15 per cent live within Cs areas, occupying 3 per cent of the land, where Cf and Cs are parts of the C category (Note 16.A).

on a cause of climate, not an aspect, a description or a symptom. Nevertheless, it is little used, because the definition of air masses is somewhat arbitrary.

Air-mass Frequency as Criterion

The climate of a location is the aggregate of the kinds of weather that prevail. So Tor Bergeron proposed in 1928 that places be classified according to the frequency with which they are affected by various air masses (Section 13.2). This is exemplified in Figure 16.3, which shows, for example, that Perth (in Western Australia) is governed by maritime polar air masses from June through September. The logical advantage of this method of classification is that it is based

Agglomerative Classification

Nowadays there are sophisticated statistical procedures for grouping places with similar climates. One is 'clusteranalysis', based on the numerical differences between climatic variables at different places, such as the annual extreme temperatures, seasonal rainfalls, frequency of hail, etc. Dozens of such climatic characteristics of each place may be considered. Classifying then involves grouping the places to minimise differences between the variables of any pair of

0 100 200 300 400

monthly total rainfall: mm

0 100 200 300 400

monthly total rainfall: mm

Figure 16.3 Thermohyet loops for Darwin, Perth and Melbourne, showing principal air masses influencing climates in each month.

them within each group. The arbitrariness of the selecting of which climatic variables to consider can be reduced but not avoided by another procedure called principal component analysis, which has been used widely since about 1980. The groupings that result depend on which procedure is adopted.

Problems

Classifying climates has several of the problems common to the whole science of classification, taxonomy. They include the following:

1 It is necessary to make an arbitrary decision about the number of categories to use. Too few and there are more frequent problems in allocating odd cases. Too many and you lose the advantage of simplification.

2 There is a suggestion of abrupt differences at the lines on a map defining the zone of each group. A way round this is to express similarity as a number, rather than in terms of yes/no criteria (Note 16.B).

3 Any hierarchical system (like that of Carl von Linne in 1735 for classifying plants, and that of Howard for clouds—see Section 8.3, and Koppen's climate classification) involves judging which attributes are of first rank of importance and which of second, etc.

4 No allowance is made for the variability and extremes which are important in many fields.

5 Insufficient distinction is made between the mesoclimate sampled by a climate station and the microclimate which affects crops and people.

6 Students of climatology often mistake allotting a class label to a climate as somehow explaining it. Koppen's classification, for instance, is merely descriptive, not an explanation.

Homoclimes

The problems just listed make some looseness unavoidable when matching homoclimes. Places within the same class do not have precisely the same climate. Nevertheless, it is at least interesting that central Chile has some resemblance climatically to the area around Cape Town in South Africa, and to the south-western corner of Australia, and that south-east Brazil is a homoclime of Madagascar. A knowledge of the climate at one place provides a feeling for what to expect at corresponding places in the same class.

The practical usefulness of the concept of homoclimes is evident in the following examples.

1 Beetles specialising in the burial of pellets of sheep dung in Australia were sought and found in homoclimes in southern Europe and the south-west of South Africa.

2 Correlations between conditions at fifty-four stations in Colombia and those in eleven tea-growing areas around the world, showed that five stations in Colombia have climates like those in part of Malaysia. So those five places were identified as possibly suitable for growing teas introduced from that area in Malaysia.

3 The transfer of bananas and sugar-cane from their origin in south-east Asia, and potatoes from Peru, has required an initial searching for homoclimes and then breeding to adapt the plants to the new circumstances.

4 Rubber originated in the Amazon valley and was prevented from spreading naturally by the quite different climates which surround it. Now it flourishes in homoclimes in south-east Asia where the daily mean temperature remains a little above 24°C or so, and the rainfall exceeds 1,750 mm/a.

In view of these examples, one concludes that climate classification is worth while, despite the difficulties.

Was this article helpful?

0 0
Solar Power

Solar Power

Start Saving On Your Electricity Bills Using The Power of the Sun And Other Natural Resources!

Get My Free Ebook


Post a comment