Introduction

The climate modelling is based on the type, the use and the shape of the studied area. At the urban scale, the type of land surface (pervious, impervious), the shape and the distribution of the buildings and the streets as well as the building use are the determinant parameters affecting the urban climate. Thus it is necessary to described accurately the urban fabric in order to apply the right energy balance.

Geoclimate is a Groovy library that implements algorithms to compute geospatial indicators (e.g. density of building, skyview factor, building compactness, road distance, …) based on vector GIS layers.

Geoclimate performs indicator computation at three spatial units.

A spatial unit corresponds to a geometry area (POLYGON or/and MULTIPOLYGON). It qualifies the best representing spatial object to compile geographical properties and characteristics. The spatial units stands for a nested spatial relationship where the building is the lowest common feature denominator as described in Bocher et al, 2018.

• the building, a collection of features that represent structures with a roof and walls, such as a house or factory,
• the block : a set of buildings that touches (at least one point in common),
• the Reference Spatial Unit, also called RSU, which is a continuous and homogeneous way to divide the space, using topographic constraint such as roads, railways, vegetation and water areas in addition to administrative boundaries.

More than 100 urban indicators are yet available. At a first stage, those indicators have been selected:

• to feed the TEB climate model developed by Météo France,
• to classify the urban tissues and build the Local Climate Zones (LCZ).

Even if Geoclimate has been developed for climate studies, the indicators can be used individually for other topics such as landscape ecology, land use, habitat conservation planning or any environmental or territory applications.