Blog · Agriculture
Microclimate: why the nearest weather station is not your field
A microclimate is the set of weather conditions in a small area, such as one field, one slope or one hollow, and it can differ noticeably from the readings at a station a few kilometres away. The nearest official station tells you about the weather at that station, measured in a standard way, and your field may be colder at night, windier, wetter or drier than that.
This matters because many farm decisions depend on thresholds. A degree or two decides whether there is frost. A short shower decides whether a field can be worked. The closer your information is to the field, the fewer surprises you get.
What a weather station actually measures
Official stations are set up to be comparable with each other, not to represent any particular farm. Under the usual meteorological conventions, air temperature is measured in a shaded, ventilated screen between about 1.25 and 2 metres above short grass, and wind is measured on a mast at 10 metres in open ground. Many stations are at airports, where the ground is flat and open.
That is a sensible standard for forecasting and climate records. It is not the same as the conditions at crop height, in a sheltered valley field, on a north-facing slope or beside a river.
Why your field differs
Several things shape the weather at field scale.
- Elevation. Air generally gets cooler with height, so a field well above the station tends to run cooler by day. At night the pattern can reverse, which is the next point.
- Cold air drainage. On calm, clear nights the ground cools quickly and chills the air touching it. That cold air is dense. It flows downhill like a slow liquid and collects in low spots and behind barriers such as hedges and embankments. A hollow can be several degrees colder than a nearby slope. Our guide to frost risk covers this in more detail.
- Slope and aspect. A slope facing the sun warms sooner and dries faster than one facing away. In the northern hemisphere that means south-facing slopes are warmer. In the southern hemisphere it is the north-facing ones.
- Shelter and exposure. Trees, hedges and buildings cut wind speed on their sheltered side. That reduces evaporation and can raise daytime temperatures, but still air also cools more at night.
- Water. Lakes, rivers and the sea moderate temperature swings nearby and raise humidity.
- Soil and surface. Dark, moist, firm soil absorbs and stores more heat by day and gives more back at night than dry, loose or mulched soil. A bare field and a field under a cover crop behave differently.
- Height above ground. On a still night the air at crop height can be clearly colder than the air at screen height. A forecast minimum of 2 °C at screen height does not rule out frost on the leaves.
Which variables differ most
Not every measurement varies to the same degree over short distances.
| Variable | How much it varies locally |
|---|---|
| Rainfall | A great deal, especially with showers and thunderstorms. One farm can be soaked while the next stays dry. |
| Night minimum temperature | A lot on calm, clear nights because of cold air drainage. Much less on windy or cloudy nights. |
| Wind | Strongly affected by terrain, shelter and height above ground. |
| Daytime maximum temperature | Usually fairly consistent across an area with similar elevation. |
| Air pressure | Hardly at all at farm scale. |
The practical lesson is that a distant station is most trustworthy for daytime temperature in well-mixed, breezy weather, and least trustworthy for rain and for calm-night minimums. Those happen to be the readings that drive frost, spray and field-work decisions.
Ways to get closer to field-level weather
There are three broad options, and they work best together.
An on-farm station. A station of your own measures your conditions directly. Its value depends on siting and upkeep. A temperature sensor in direct sun, a rain gauge under a tree, or a wind sensor behind a shed will mislead you. Sensors drift and rain gauges block, so they need checking. One station also describes one spot, which may not be the frost pocket at the bottom of the farm.
Gridded weather data. Weather services and data providers produce gridded estimates, where observations, radar, satellite and weather models are combined to estimate conditions for every cell on a map. This gives a value for every field, including ones with no sensor. The limits are that each cell is an average over its area, fine terrain effects can be smoothed out, and the estimate is only as good as the inputs and the model.
Local knowledge. You probably already know which field holds frost, which dries last and where the wind funnels. That knowledge is real data. Writing it down against field names makes it usable by other people and by software.
A cheap way to test any source is to compare it with reality. Put a simple minimum thermometer in the suspected cold spot for a few calm nights, or compare a gauge in the field with the reported rainfall after a showery week. If the source is consistently off by a similar amount, you can allow for it. If it is erratic, treat it with caution.
What field-level weather feeds into
Field-level weather is rarely the end goal. It is the input to other calculations:
- Growing degree days, which track crop and pest development from daily temperatures.
- Frost warnings, which depend on the minimum at crop height in the coldest part of the field.
- Spray windows, which depend on wind, temperature, humidity and inversions at the time and place of spraying.
- Field-work windows, which depend on recent rain and drying conditions.
Small errors in the input become larger errors over a season. A temperature bias of one degree, added up daily, moves a growing degree day total by a meaningful amount. That is a good reason to know where your numbers come from.
Weather data informs a decision. It does not make it. Product labels, your own observations in the field and advice from your agronomist or local extension service take precedence over any figure on a screen.
Where Prism fits
Prism Microclimate, built with FarmGrid, covers 961k fields across one whole state. It provides field-level weather, soil moisture, frost, spray and field-work windows, growing degree days and a morning brief per field. You can read more on our agriculture page, or see the other industries we work in.