Loading…
Generating future weather files under climate change scenarios to support building energy simulation – A machine learning approach
•A workflow to generate weather files under four climate change scenarios is proposed.•Quantile-Quantile bias-correction is used to remove existing bias in GCM.•A hybrid model of k-nearest-neighbour classification and Random Forest regression is used to downscale GCM data.•Future weather files, year...
Saved in:
| Published in: | Energy and buildings 2021-01, Vol.230, p.110543, Article 110543 |
|---|---|
| Main Authors: | , , |
| Format: | Article |
| Language: | English |
| Subjects: | |
| Citations: | Items that this one cites Items that cite this one |
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | •A workflow to generate weather files under four climate change scenarios is proposed.•Quantile-Quantile bias-correction is used to remove existing bias in GCM.•A hybrid model of k-nearest-neighbour classification and Random Forest regression is used to downscale GCM data.•Future weather files, year by year, under different climate change scenarios are constructed.
General circulation models (GCM) have been used by researchers to assess the effect of climate change in different fields of study. In the case of building energy performance, GCMs can be used to evaluate future building energy performance through simulations. However, a key issue with the use of GCM data in building energy simulation is the inadequate resolution and bias of the data. Therefore, in order to use this data for simulation purposes and better predict future building performance, further processing is required. The first challenge is that the GCMs are usually biased, which means a considerable deviation can be found when the historical GCM data is compared to station observed weather data. The second challenge is that the GCM data has daily temporal resolution rather than the hourly resolution required in building energy simulation.
In order to utilize GCM data to estimate future building performance through simulation, the current study suggests a workflow that can be applied to climate change data. First, a bias-correction technique, known as the quantile-quantile method, is applied to remove the bias in the data in order to adapt GCMs to a specific location. The study then uses a hybrid classification-regression model to downscale the bias-corrected GCM data to generate future weather data at an hourly resolution for building energy simulation. In this case, the hybrid model is structured as a combined model, where a classification model serves as the main model together with an auxiliary regression model for cases when data is beyond the range of observed values. The proposed workflow uses observed weather data to determine similar weather patterns from historical data and use it to generate future weather data, contrary to previous studies, which use artificially generated data. However, in cases where the future GCM data showed temperatures ranging outside of the observed data, the study applied a trained regression model to generate hourly weather data.
The proposed workflow enables users to generate future weather files year by year under different climate change scenarios an |
|---|---|
| ISSN: | 0378-7788 1872-6178 |
| DOI: | 10.1016/j.enbuild.2020.110543 |