Microclimate modelling

Why model microclimate?

Temperature and water affect all living things. They influence every aspect of the physical environment and govern any process that involves energy flow, setting boundaries on what organisms can or cannot do. Most studies of climate biology rely on data derived from widely-spaced weather stations. Such data mask spatial variability in microclimate and ultimately bear little resemblance to the conditions that organisms experience in the wild.

Microclimate modelled at 10-minute intervals in a Sheltered Valley in Devon, UK in May 2015.

The differences between reference temperatures (measure by weather stations) and microclimate temperatures can be dramatic.

The temperatures between areas of a 20 cm x 30 cm south-west facing slope in Cornwall have a greater range than temperatures across the whole of the UK.

20 x 30 cm of a SW-facing slope in CornwallRange: 21.9 ºC
UK temperature (2013-05-22 14:00)Range: 18.67 ºC

The microclimate research group are pioneering the development of microclimate models and their application in conservation and agricultural research. This has included the development of software and tools for modelling microclimate and fine-scale hydrology and work demonstrating that microclimate significantly buffers ecological responses to climate change, alters crop-pest disease risk and dictates where high-value crops can be grown.

Our models accurately predict temperature and soil moisture at sub-metre resolutions. These can be applied over regional extents.

Thermodynamic principles are used to capture the effects of terrain, sea-surface temperature, altitude and surface albedo on local temperature and water availability.

High spatial, but low temporal resolution remotely-derived landscape data and low spatial, but high temporal resolution weather data are used to drive the models.


The functions in the microclima package contain tools for modelling the mechanistic processes that govern fine-scale variation in temperature. It includes tools for determining local variation in temperature arising from variation in radiation, wind speed, altitude, surface albedo, coastal influences and cold air drainage. A series of functions for determining the fine-scale topographic and vegetation effects on wind speed and radiation are also provided. It also includes tools for deriving canopy cover, leaf architecture, surface albedo and cold air drainage basins from digital elevation data and aerial imagery.

The R package can be installed from https://github.com/ilyamaclean/microclima

Function help files and a guide to running the models are available here.


The functions in the ecoyhdrotools package contain tools for modelling fine-scale hydrological processes. It includes tools for delineating basins, calculating flow accumulation and velocity and for calculating evapotranspiration at daily our hourly intervals, using high resolution microclimate sand digital elevation data as inputs. It also includes functions for downscaling rainfall and generating synthetic series of sub-daily rainfall, consistent with daily totals, by applying the Bartlett_lewis rectangular pulse model.

The package can be installed from https://github.com/ilyamaclean/ecohydrotools

*NB This package is currently under development. Some functions have not been fully tested and additional functionality will be added shortly

We are also working on integrating microclima with the NicheMapR package described here.

This will enable fine-resolution grids of microclimate to be generated for anywhere in the world in a fully automated way. Watch this space.

Our models couple meteorological data with fine-grained environmental captured remotely. Remote-sensing technology has now evolved to the point where the 3D structure of landscapes and vegetation communities can be mapped with relative ease. The fusion of multispectral imaging capabilities with fine-grained structure-from motion photogrammetry techniques, all of which we can capture ourselves from drone platforms creates new possibilities for mapping and modelling the physical determinants of microclimate such as foliage structure and terrain.

We are working with the DroneLab research group at the University of Exeter's Environment and Sustainability Institute.

Drone photos taken from many different angles can be stitched together to create 3D models of landscapes and gives much greater clarification of canopy shading and microtopography of an area.

Example outputs of our microclimate models (2pm, 23rd May 2010)

Key references

Maclean IMD, Mosedale JM, Bennie JJ (2018) Microclima: An R package for modelling meso- and microclimate. Methods Ecol Evol. 2018;00:1–11. https://doi.org/10.1111/2041-210X.13093

Maclean IMD, Suggitt AJ, Wilson RJ, Duffy JP & Bennie JJ (2017) Fine‐scale climate change: modelling spatial variation in biologically meaningful rates of warming. Global Change Biology 23(1): 256-268

Maclean IMD, Bennie JJ, Scott AJ & Wilson RJ (2012) A high-resolution model of soil and surface water conditions. Ecological Modelling 237–238: 109-119