Wind atlases no substitute for on-ground experience

Atlases and the maps they contain have been used for hundreds of years as a means of navigating the unknown. Wind atlases serve the same purpose. They are designed to help stakeholders identify locations of high wind speeds and understand the overall potential for wind development in an area.

The first wind maps were developed in the early 1980s, culminating in one of the best-known examples, the European Wind Atlas, in 1989.

Since then, more than 100 countries have been mapped at varying levels of detail. The work is usually driven by a desire to kick-start the wind industry in any given region.

The work has generally moved out of research institutes into the hands of commercial enterprises, which are continually developing more advanced tools and skills sets.

The generation of a wind atlas is usually the first step towards defining national wind targets and development areas.

It can act as a focal point for the development of a wind-power industry, by convincing governments, investors and private industry that there is an opportunity to be exploited.

It is not surprising to see many developing markets proceed down the wind-atlas route, with industry bodies and development banks keen to help them.

The International Renewable Energy Agency and the World Bank have sponsored and promoted large-scale wind mapping exercises, focused mainly in developing markets.

Yet, even Europe, which has some of the best understood wind resources in the world, is being reinvestigated based on a uniform mesoscale-to-microscale model-chain methodology at high resolution, under the New European Wind Atlas (NEWA) project.

Wind atlas modelling takes a top-down approach, by creating high-level simulations at coarse resolution for large areas and then moving to higher resolutions within specific geographical bounds.

Historically, the first wind maps were generated using wind measurements from several sites in a specific region. Of course, this approach only worked if sufficient quantities of high-level data were available, limiting the creation of wind maps to areas with well documented wind regimes.

Numerical wind atlases were created to get around this problem. The most common method is to use a mesoscale model to simulate the wind field on a regional basis, based on inputs from global climate parameter datasets.

This model is then combined with microscale wind-flow models to downscale the map to a higher resolution.

Mesocale models

The two most commonly used mesoscale models are the Karlsruhe Atmospheric Mesoscale Model (Kamm) and the Weather Research and Forecasting model (WRF).

Although Kamm can be more easily coupled with microscale-flow models like the Wind Atlas Analysis and Application Program (Wasp), which requires less computing power and is quicker to run, this combination struggles with accuracy in complex geographical and atmospheric conditions, such as coastal areas.

WRF can use more up-to-date climate datasets as inputs, model with greater temporal granularity and, when combined with a computational fluid dynamics model, can better capture flow over complex terrain and can be tuned for different atmospheric conditions.

In the future, more advanced modelling techniques, such as those using a large eddy simulation model, will allow detailed flow maps at a much higher resolution (<1km) to be created.

However, this technique is even more computationally intensive and would be difficult to run for long-term simulations. There are also some known issues with the results that will need to be addressed before we see widespread adoption.

In many cases, it might be more immediately beneficial to make current and older models publicly available, so that they can be used as an input to developers’ microscale models, which is what the NEWA project is aiming to do.

It’s not to say that current wind atlases have all the answers. Any atlas is only as good as the information used to create it. All maps have uncertainties associated with their accuracy and wind maps are no different.

An important step is to try and use large amounts of measured data to validate numerical wind maps. Even then, the sheer scale of modelling requires assumptions to be made, which then introduce uncertainties.

There are other potential wildcards to consider, ranging from certain microclimates that are hard to model through to the future effects of climate change.

The biggest issue with wind maps seems to be a misunderstanding of their limitations by users, leading to bad decision-making.

A wind atlas is a detailed guide, but it cannot replace actual experience gained on the ground. Wind atlases are useful tools but there are still parts of the map that should read "here be dragons".

Ioannis Papadopoulos is business lead for renewable energy analytics at DNV GL