Weather - particularly in terms of cloud formation and precipitation - can have a large influence on a planet's radiative balance. The reason is that clouds and snow cover both alter the albedo of the planetary surface beneath. If a rock surface is beneath a thick cloud cover, the clouds can reflect much radiation back into space before it ever reaches the relatively low albedo surface. The same is true for snow - fresh snow has a fairly high albedo of about 0.8, which means that typically snow-covered terrain is much more reflective than bare terrain. Both effects would usually tend to cool the planet.

However, clouds also have a fairly high IR opacity - low, dense clouds can hold up to 60% of the outgoing IR radiation, high Cirrus clouds still 10% - so cloudcover prevents the terrain from cooling rapidly at night, and this effect works to keep a planet warmer.

It should be clear that when we talk about weather at this stage, this refers to a planet that is Earth-like, i.e. has liquid water as the medium creating clouds and weather patterns. This need not be the case - worlds significantly colder (such as Titan in the Solar system) might have a different medium such as methane that can condense and form rain and lakes on the surface. Such environments will be covered later.

The weather simulation consideres four different types of clouds - low-level convective and synoptic weather pattern clouds, mid-level clouds and high level Cirrus clouds.

Most information I have on synoptic weather patterns, jet streams, cloud formation, IR opacity of clouds and global circulation is from Meteorology for Scientists and Engineers by Roland B. Stull

Convective clouds

When the sun heats the surface of a planet, it also heats the air in contact with that surface. Since that air parcel is now warmer than its surroundings, it starts to rise. Especially if the ground is moist, it carries a lot of humidity with it. When this air parcel rises, it also expands and cools, and at some point the humidity condenses into small droplets - a visible Cumulus-type cloud is formed.

Convective cloud development depends on the heating of the ground, thus it usually starts up in the late morning when the ground warms, reaches its maximum in the afternoon and cloud cover decays towards evening when the ground cools again. Strong convection may lead to the formation of thunderstorms and copious rain. Also, the mechanism is particularly pronounced over low albedo terrain (because this leads to quick heating of the ground) and more effective over land than over the ocean. Regions with globally sinking air - such as the subtropical ridges - can suppress convective cloud formation.

In terms of the effect of temperature, convective clouds have a cooling effect, because they start to appear precisely when the terrain gets heated and lower the radiation flux onto the ground, while they dissolve at night which allows the longwave IR radiation to cool the surface again.

In the simulation, convection is automatically active whenever weather is declared at all.

Synoptic clouds

The fact that the axial tilt of Earth is such that predominantly the equatorial region is heated and the polar regions stay cooler gives rise to the formation of meandering high-altitude jet-streams at mid-latitudes. These in turn drive the formation of cyclones, low-pressure weather systems at low altitude. Each such system corresponds to dense low-level clouds as well as a core of precipitation.

The simulation automatically generates synoptic weather patterns by continuously adding and removing a random selection of cyclones in the jet stream latitudes. Note however that this precise weather pattern is really contingent on the axial tilt being low - if the polar regions receive most radiation (as for a high axial tilt), weather patterns will be different - just as they will be different when the axial rotation period of the planet in much faster or slower than 24 hours.

Mid- and high-level clouds

Processes leading to the formation of mid and high-level clouds are generally complicated. Partially these clouds are driven by synoptic weather, partially not. Since both layers have different albedo and IR opacity when compared with low-level clouds, they are nevertheless included in a probabilistic way.

The simulation allows to specify their minimal and maximal coverage, and every eight hours a new random value is determined for the surface element in question.

For mid-level clouds the value represents the 100% opacity equivalent - so a visible sky covered in a thin mid-level Altostratus layer is not represented by a fraction of 1.0 because the cloud itself is not opaque. High level Cirrus clouds are assumed to be 50% opaque, so even a coverage fraction of 1.0 allows light to pass through to the ground.

The following is a sample weather declaration for on average 20% mid- and high-level cloud cover. This automatically activates also the simulation of convection and Earth-like synoptic weather patterns.

mid_cloudlevel_min 0.0
mid_cloudlevel_max 0.4
high_cloudlevel_min 0.0
high_cloudlevel_max 0.4

It is possible to visualize the current clouds (selected by altitude) in a 2-d plot using the keywords cloud_cover or differential in laters as low_cloud_cover, mid_cloud_cover and high_cloud_cover.

file earth_clouds.dat
var_z low_cloud_cover

The result might look as follows with three small systems in the Northern and two large cyclones in the Southern hemisphere:

Cyclone systems visualized as region of low clouds on Earth

After these preparations, we're now ready to introduce and study a simple model of Earth.

Continue with Earth.

Back to main index     Back to science     Back to worldbuilder

Created by Thorsten Renk 2023 - see the disclaimer, privacy statement and contact information.