# Climate, Biomes and Environmental Variation

> Rain falls where moist air rises and cools and is scarce where air sinks and warms, and the warmer a month, the more rain it needs to count as humid; read month by month, warmth and water give a first guess at a place's biome. This lesson builds that rule from sunlight, the circulation of air and ocean, and mountains, reads it on a Walter-Lieth climate diagram and by Köppen-Geiger thresholds, and then tests it on a made-up savanna-climate station, set beside a tropical analysis in which forest and savanna both occur across similar climates and fire may help decide between them. It closes with what climate alone cannot tell you: soils, the timing of rain, microclimate, and the difference between observed and projected change.

Course: Ecology and Evolution (Environments and populations). Level: core university (third year). About 20 minutes.
Lesson page (link this one when citing): https://stemhop.com/lessons/climate-and-biomes/

## What you'll learn

- Explain that temperature, rainfall and their seasonality largely organize where biomes occur, and that fire, soils, rainfall timing and microclimate modify the pattern.
- Explain how latitude sets solar input per square metre, why axial tilt rather than distance makes the seasons, and how uneven heating drives the atmospheric circulation, idealized as Hadley, Ferrel and polar cells, with rising moist air near the equator and sinking dry air in the subtropics.
- Explain how winds and Coriolis deflection drive surface ocean currents that carry heat poleward, and why mountains have a wet windward side and a dry leeward side.
- Read a Walter-Lieth climate diagram (10 degrees C to 20 mm), identify dry periods, classify a station by Köppen-Geiger thresholds, and state what climate cannot predict about a biome (fire, soils, rainfall timing, microclimate).
- Distinguish weather from climate (30-year normals), and observations from scenario projections.

A **biome** is a large region with similar climate and similar dominant vegetation and animals: tropical wet forest, savanna (grass with scattered trees), desert, temperate grassland, chaparral, boreal forest, tundra.

> **The one idea.** Rain falls where moist air rises and cools, and is scarce where air sinks and warms. The warmer a month, the more rain it needs to count as humid: on a Walter-Lieth climate diagram a month is read as dry when its rain, in millimetres, is less than twice its temperature in degrees C. Read month by month, warmth and water give a first guess at the biome.
>
> *Scope:* broad biomes, from 30-year averages; not where any one species lives. The air rule is a broad-scale tendency: rising air rains only if it carries enough water vapour, and fronts and mountains make rain too. Fire, soils, the timing of rain and local microclimate can overturn the biome guess (second half of the lesson).

**Weather** is the state of the atmosphere at a given time and place; **climate** is its long-term pattern, summarized by convention as 30-year averages called climate normals.

## Warmth: over a year, sunlight per square metre falls with latitude

The first film’s tilted torch has a number. The fraction of sunlight that reaches a horizontal surface is the cosine of the **solar zenith angle**, the angle between the Sun and the point directly overhead. With the Sun straight overhead, cos 0° = 1: all of it arrives. At 60 degrees from overhead, cos 60° = 0.5: the oval of light is twice as long, so each patch of ground gets half as much.

> **energy on a horizontal surface = energy on a surface facing the Sun x cos(zenith angle)**

**Worked problem A.** At the **top of the atmosphere**, before air and cloud absorb or reflect any of it, a surface facing the Sun receives about 1,360 W m−2 (watts per square metre). With the Sun 60 degrees from overhead (roughly noon at latitude 60 degrees at an equinox), a horizontal surface receives 1,360 x cos 60 = 1,360 x 0.5 = **680 W m−2**.

Day length matters too. At a **solstice** (June, December) one hemisphere leans furthest toward the Sun; at an **equinox** (March, September) day and night are about equal at most latitudes. The table’s point: on a midsummer day, long days at high latitude beat the slant; over a year they do not. Daily means at the top of the atmosphere, W m−2, for a tilt of 23.44 degrees and a constant Earth-Sun distance (computed):

| Latitude | Year-round mean | June solstice day | December solstice day |
| --- | --- | --- | --- |
| Equator | 415 | 397 | 397 |
| 30° N | 365 | 490 | 220 |
| 60° N | 236 | 492 | 24 |
| North Pole | 172 | 541 | 0 |

Over a year the tropics receive about 2.4 times what the poles do. That imbalance drives the air: the tropics gain more energy than they radiate away and the poles lose more, so the atmosphere and oceans carry heat poleward.

**Predict.** Earth is closest to the Sun in early January. If distance made the seasons, what would the two hemispheres both have in January? What actually differs between them?

> **Check your prediction.** Both would have summer at once; instead their seasons are opposite. Earth’s axis is tilted about 23.4 degrees from the perpendicular to the plane of its orbit, so each hemisphere in turn leans toward the Sun, getting more direct light and longer days. That, not distance, makes the seasons. Distance changes sunlight by only about 7% between early January and early July; at 45 degrees north a midwinter day still gets only about a quarter of a midsummer day’s sunlight, distance included.

## Water: where air rises and where it sinks

The first film’s sinking air dries because of a quantity it did not name: **relative humidity**, the water vapour present as a share of the amount that would saturate air at that temperature. Sinking air keeps its vapour but warms, so its relative humidity falls and cloud does not form. The Kalahari, like the Sahara, lies under this subsiding air.

The loop is the **Hadley cell**; at the surface, air flows from the subtropics back toward the equator. The idealized model adds a **Ferrel cell** in mid-latitudes and a **polar cell** at high latitudes. It is a cartoon: the real circulation is full of waves and eddies, and only the Hadley cell is driven directly by heating; the Ferrel cell is the averaged effect of mid-latitude storms.

The film’s swinging rain belt sets each place’s dry season, from none at some places near the equator to most of the year at its edges, and land and sea bend its path.

Earth’s rotation, through the **Coriolis effect**, deflects moving air and water to the right of their motion in the Northern Hemisphere and to the left in the Southern. It matters over hundreds to thousands of kilometres; it does not decide which way a sink drains.

**Predict.** In the Northern Hemisphere that surface flow heads south. Turned to the right, from which direction does it blow?

> **Check your prediction.** Turning right from a southward heading points the air toward the west, so the wind blows from the northeast. These are the easterly **trade winds**. Air moving poleward in mid-latitudes is turned toward the east and becomes the **westerlies**.

## Oceans and mountains bend the pattern

Winds, turned by Coriolis deflection, drive large surface currents that carry warm water from the tropics poleward and return cold water toward the equator. Density differences, set by temperature and salinity, help form deep water, which with winds, eddies and mixing sustains a slower deep circulation. Coasts beside warm currents are milder than their latitude suggests, and coasts beside cold currents are often cool and dry.

The popular claim that the Gulf Stream “keeps Europe warm” overstates the case: much of the winter contrast between western Europe and eastern North America comes from the atmosphere, although the ocean still moves a large share of heat poleward.

Mountains apply the same rule locally: the **windward** slope gets the rain, and the **leeward** side lies in a dry **rain shadow**.

*Figure (AI-generated illustration).* Moist air forced up a slope cools and drops its rain on the windward side; descending the far side it warms and dries, leaving a rain shadow. On mountainous islands such as Hawaii the windward slopes facing the trade winds are wet and the leeward sides much drier. Image: side view of a mountain ridge rising from the sea. Moist air from the left rises up the windward slope, where a cloud rains on dense green forest. On the far slope a single gold arrow descends over sparse pale vegetation with no cloud. A small two-step inset shows air cooling as it rises and warming as it sinks.

## Reading a climate diagram

Why does the film’s **Walter-Lieth diagram** draw the dry line at twice the temperature? The ratio is empirical, a marker of months dry enough to stress plants, not a measured water balance: **potential evaporation**, the water a wet surface would lose to the air, can be well above the line in a warm month (Going deeper). Always read the axis labels, the station, its elevation and the years averaged.

The **Köppen-Geiger** classification uses fixed thresholds and a letter code. It asks first whether a place gets too little rain for its warmth: it is arid (**B**) when its annual rain in mm is below 10 times a threshold that rises with mean annual temperature. In Beck and colleagues’ (2018) version, with MAT the mean annual temperature, the threshold is 2 x MAT + 28 when more than 70% of the rain falls in the warmer six months (April to September or October to March), 2 x MAT when more than 70% falls in the colder six, and 2 x MAT + 14 otherwise. A station that is not arid gets a temperature class, such as **A** (tropical: every month at least 18 degrees C), then that class’s own subtype rules. For class A the driest month decides: **f**, no dry month (at least 60 mm); **m**, monsoon; **w**, a marked dry season.

**Worked problem B: station A.** In this made-up data set, a lowland station has a mean temperature of 26 degrees C in every month. Mean monthly precipitation, January to December, is 140, 120, 150, 180, 200, 160, 140, 110, 90, 100, 130 and 160 mm. Assume these are 30-year means.

1. *Annual total.* The twelve months sum to **1,680 mm**.
2. *Dry period?* At 26 degrees C the red line sits at 52 mm; the driest month, September, has 90 mm. Every month is humid: **no dry period**.
3. *Köppen-Geiger class.* April to September holds 880 mm (52.4%), October to March 800 mm (47.6%). Neither half is warmer or holds more than 70%, so the “otherwise” threshold applies: 2 x 26 + 14 = 66, an arid cut-off of 660 mm, well below 1,680 mm: not arid. Coldest month 26 degrees C, at least 18: class A. Driest month 90 mm, at least 60: **Af, tropical rainforest climate**.
4. *Biome.* Weak temperature seasonality and year-round rain fit tropical wet forest. That is a hypothesis to test against the vegetation present.

*Interactive on the lesson page:* Climate diagram and class — an invented station

Keep station A’s average warmth, cut its rain by a third and crowd most of it into seven months, and the second film shows what the diagram says, and what it cannot.

**Film transcript: Same Warmth, Two Answers**

### Film transcript: Same Warmth, Two Answers

The first film built a rule. Rain falls where moist air rises, and warmer months need more rain to count as humid. Now test it on two made-up stations. Both average about 26 °C. What grows at each? Station A is the one you just worked. Its blue rain line rides above the red in every month. No month is dry. Its class is Af: tropical, with no dry month. That is a tropical rainforest climate. Station B has the same average warmth, though its months run from about 24 to 30 °C. But it gets about two thirds as much rain, and most of it falls from April to October. From November to March, the blue line drops below the red. Those five months fill with gold.

They are the dry season. Its class is Aw: tropical, with a marked dry season. That is a tropical savanna climate. The rule's first guess is grass with scattered trees. So far the rule gives two first guesses. Similar warmth, different rain, different climate classes. Now look at station B's rain once more. Staver, Archibald and Levin analysed tree cover across the tropics. They looked at places with about 1,000 to 2 ,000 and a half -thousand millimeters of rain a year. Those places also had a dry season shorter than seven months, by the study's own measure. There, tree cover was rarely in between. Places were mostly savanna, or mostly forest.

Station B's rain total falls in that range, and its diagram points to savanna. Would every place with a climate like that be savanna? No. In that window, savanna and forest were both found under similar climates. And fire was associated with keeping areas open. Where grass is open, fire can keep it open. A closed canopy can hold fire back. So each state may help keep itself going.

This is a pattern, not proof. Savanna and forest may be alternative states under similar climates. Whether those states are truly stable is still debated. So climate gives a first guess at the biome. Fire, soils, the timing of rain and microclimate can change the answer. A savanna climate does not guarantee a savanna.

From the film

- Two places with similar average warmth can get different first guesses at the biome because of how much rain falls and when.
- Station A has no dry month and a tropical rainforest climate (Af); station B has five dry months on its climate diagram and a tropical savanna climate (Aw).
- In Staver and colleagues' analysis, at about 1,000 to 2,500 mm of rain a year with a dry season shorter than seven months (by their own measure), tropical tree cover was mostly either savanna-like or forest-like, rarely in between.
- Fire was associated with keeping savanna open; the proposed feedback is that fire can keep open areas open and a closed canopy can hold fire back, so the two may be alternative states under similar climates.
- A savanna climate does not guarantee a savanna: climate gives a first guess, and fire, soils, rain timing and microclimate can change the answer.

## Biomes are more than climate

The second film gave station B’s class without the arithmetic. Its temperatures, January to December, are 24, 26, 28, 30, 29, 27, 26, 25, 26, 27, 26 and 24 degrees C; its rain 5, 10, 30, 80, 150, 180, 200, 220, 160, 60, 15 and 5 mm.

Total 1,115 mm, mean 26.5 degrees C. April to September, the warmer half, holds 88.8% of the rain, so the threshold is 2 x 26.5 + 28 = 81 and the arid cut-off 810 mm: not arid. Coldest month 24 degrees C: class A. Driest month 5 mm, below 60 and below the monsoon (Am) cut of 100 − (annual total / 25) = 55.4 mm: **Aw, tropical savanna climate**.

*Fire and alternative states.* The film’s window comes from Staver, Archibald and Levin’s (2011) analysis of tree cover in tropical Africa, Australia and South America. Ecologists call the film’s pattern, tree cover rarely in between, **bimodal**. The study is correlational: it shows what occurs together, not cause, so, as in [how to test an evolutionary explanation](https://stemhop.com/lessons/testing-evolutionary-explanations/), keep pattern and mechanism apart. The study measured dry-season length its own way, not by the Walter-Lieth count, so station B’s 1,115 mm total falls within the study’s rainfall range, but its five Walter-Lieth dry months do not show that it lies inside the full window. The strength of the film’s fire feedback depends on local conditions. Whether these states are truly stable is still debated.

*Soils.* Water-holding capacity and nutrients differ under the same rain, so the same climate can support different vegetation.

*Rainfall timing.* Annual totals hide timing: a few large storms can soak deeper or run off, and long gaps can dry the topsoil. Whether timing alone changes the plants is a question for an experiment.

### Design problem

Two grasslands receive the same mean annual rainfall, but at one it comes in frequent small showers and at the other in a few large storms. Design an experiment to test whether rainfall timing, with the total held constant, changes plant production.

> **One way to do it.** Comparing the two sites cannot separate timing from soils, species pools and history, so manipulate timing within one grassland. Build rain-out shelters over randomly assigned, replicated plots and reapply the rain by irrigation: the natural pattern on some, the same total in fewer, larger events on others. Unsheltered plots check that the shelters themselves (shade, warming) do not change the result. Log soil moisture to confirm the treatment made it more variable, and measure aboveground net primary production (new plant growth above the soil) and species diversity for several years, since one year’s weather can dominate. Knapp and colleagues (2002) did this for four years in a native mesic grassland, one with moderate rather than scarce rainfall. Fewer, larger events raised soil-moisture variability, lowered production and raised species diversity. The answer applies to that grassland.
>
> *Figure (AI-generated illustration).* Holding the total rain constant while changing only the size and spacing of storms separates rainfall timing from rainfall amount. In Knapp and colleagues' grassland, fewer, larger storms lowered plant production; repeat it elsewhere before generalizing. Image: three grassland plots in a row. The left plot is open to natural rain. The middle and right plots sit under rain-out shelters with collecting tanks and irrigation hoses; the middle one is sprinkled in many small events and the right one in two larger events, with the same total water. A small inset compares the unsheltered plot with the sheltered natural-pattern plot.

### Microclimate

**Predict.** A weather station sits in open ground, as stations normally do. A logger hangs beneath a forest canopy nearby. On a hot summer afternoon, which reads higher? In a cold spell?

> **Check your prediction.** The open station reads hotter in the heat. De Frenne and colleagues (2019) compiled 714 paired temperatures beneath forest canopies and in nearby open habitat at 98 sites on five continents: understories averaged about 1.7 degrees C cooler in mean temperature and 4.1 degrees C cooler in maximum temperature, and canopies warmed the understory in cold conditions. The temperature an organism experiences is not the station’s.

### Observation versus projection

Beck and colleagues (2023) reconstructed that about 5% of land outside Antarctica changed major Köppen-Geiger class between 1901 to 1930 and 1991 to 2020. Between 1991 to 2020 and 2071 to 2099 their models project 5% under the low-emission scenario SSP1-2.6, 8% under the middle scenario SSP2-4.5 and 13% under the high-emission scenario SSP5-8.5. The first is a reconstruction from observations; the projections are model results under stated scenarios. Neither is a biome change, which depends on dispersal, soils, fire and time.

## Going deeper

### Worked problem A, continued: the whole-planet average

Earth intercepts sunlight over a disk of area pi r2 but spreads it, through day and night, over a sphere of area 4 pi r2. The global mean is 1,360 / 4 = **340 W m−2**, a top-of-atmosphere average; about 29% of incoming sunlight is reflected back to space.

### Who explained the Gulf Stream’s share

Seager and colleagues argued in 2002 that the atmosphere carries much of the contrast: prevailing westerlies bring air warmed by heat the ocean stored over summer onto Europe’s west coast, and the large-scale wind pattern, shaped partly by the Rocky Mountains, brings cold continental air to eastern North America. Changes in Gulf Stream heat transport still affect Europe’s climate over decades.

### Climate diagrams in practice

The current World Meteorological Organization standard period for normals is 1991 to 2020. Above 100 mm per month, Walter’s own diagrams compressed the rain scale, and ClimateCharts.net keeps the 1 degree C to 2 mm ratio only up to 100 mm. Thornthwaite’s potential evaporation at a steady 26 degrees C is about 125 mm for a standard 30-day month of 12-hour days, before the day-length corrections a real station needs: over twice station A’s 52 mm line. Actual evaporation is capped by the water available.

### Plants of hot, dry biomes

In hot, dry biomes many plants use C4 or CAM photosynthesis, variations on the carbon fixation covered in [the light reactions](https://stemhop.com/lessons/light-reactions/) and [the Calvin cycle](https://stemhop.com/lessons/calvin-cycle/).

---

From STEMhop (https://stemhop.com), free lessons by Christopher Deppmann. Lesson text © the authors. Practice questions: https://stemhop.com/lessons/climate-and-biomes/
