Sun and Shadow Analysis on Terrain

Solar position — the sun's azimuth and elevation — follows from date, time of day and geographic position. It is pure astronomy and it is exact; nothing about your data affects it.

This is also why the geographic coordinates of your project matter for sun analysis even if all your measurements are in a projected grid. A site placed at the wrong longitude gets the right sun path at the wrong time of day.

For a given sun position, every cell is tested: does the terrain between this cell and the sun rise high enough to block the ray? If yes, the cell is shaded.

Two distinct effects come out of this. Self-shading, where a slope faces away from the sun and gets no direct light regardless of what is around it. And cast shadow, where a ridge or a pit wall throws shade onto ground that is otherwise well oriented.

Cast shadow is the one that surprises people. A south-facing slope in the northern hemisphere looks ideal until you notice the ridge to its east delays sunrise there by two hours every day of the year.

Repeat the shadow test through the day at a fixed time step and count the steps each cell was lit. That count, scaled, is sunlight hours — the simplest and most intuitive output.

Insolation goes further. Instead of counting whether light arrived, it accumulates how much: the intensity depends on the angle between the sun and the surface, so a steep slope catching the sun obliquely receives less energy per square metre than flat ground under the same sun.

Use hours when the question is duration — will this bench be workable by 9 am, does this orchard row get six hours. Use insolation when the question is energy — which of these two slopes will produce more from the same panels.

STREAM computes both sunlight hours and solar insolation over the terrain.

The date. There is no single "sun" result. Winter solstice is the worst case and the one that governs system design; equinox is the balanced case; summer solstice flatters everything. Analysing one date and calling it the answer is the most common mistake.

The time step. A coarse step misses narrow shadows from thin obstructions. A very fine step costs time without adding much once the step is well below the duration of the shadows you care about.

The extent. Shading comes from terrain that may sit well outside your area of interest — a mountain 5 km east still delays your sunrise. If your DEM stops at the site boundary, your mornings will look sunnier than they are.

The surface. On bare-earth terrain, buildings and trees cast no shadow. That is right for assessing the landform and wrong for assessing a specific roof.

Solar siting: ranking candidate areas by real received energy rather than by aspect alone, and finding the rows that a ridge will shade every morning.

Agriculture: frost pockets, orchard and vineyard row orientation, and which fields warm first in spring.

Mining and quarrying: when a face or bench comes into sun, which matters for ice, for thaw and for scheduling.

Site and building planning: overshadowing of neighbours, and where amenity space will actually be sunny.

Snow and hydrology: melt timing follows sun exposure closely on complex terrain.

This is a geometric model of direct sunlight on a surface. It does not know about clouds, haze, or the diffuse light that a real sky delivers even in shade — which is why a fully shaded cell is not actually at zero energy in reality.

It also does not model the panels, the trackers, the soiling or the temperature losses. Terrain sun analysis tells you what the landform does to the sunlight. Turning that into predicted yield is a separate calculation with its own assumptions.