[595] - Atmospheric Downscaling using Genetic Programming

All available metadata of the dataset is listed below. Some features are available, e.g. download of dataset or additional description file.

By downloading files from this dataset you accept the license terms of TR32DB Data policy agreement and TR32DBData Protection Statement.
Adequate reference when this dataset will be discussed or used in any publication or presentation is mandatory. In this case please contact the dataset creator.
Due to the speed of the filesystem and depending on the size of the archive and the file to be extracted, it may take up to thirty (!) minutes until a download is ready! Beware of that when confirming since you may not close the tab because otherwise, you will not get your file!
Zerenner, T., 2013. Atmospheric Downscaling using Genetic Programming. PhD Report, Meteorological Institute, University Bonn, Bonn, Germany. Accessed from https://www.tr32db.uni-koeln.de/data.php?dataID=595 at 2019-08-20.
Citation Options
Export as: Select the file format for your download.Citation style: Select the displayed citation style.
Title(s):Main Title: Atmospheric Downscaling using Genetic Programming
Description(s):Abstract: We are using a machine learning approach, Genetic Programming (GP), for discovering downscaling rules from a training data set (high-resolution model runs). This report covers our rst steps to adapting GP to our downscaling problem. Exact reproduction of the ne-scale elds from the coarse data will not be possible. We introduce multi-objective tness functions aiming at reproducing a 'realistic structure' rather than the exact eld. Furthermore we integrate functions operating in space to make the GP able to account for the spatial nature of our data in a more exible way. We show that especially the integration of spatial standard deviation into the tness function gives very promising results. In the rst part of this report the di erent modi cations of our original GP algorithm are described and tested on the problem of downscaling near-surface temperature in clear sky nights. The second part covers a rst test for downscaling of near-surface wind speed. The last part describes an approach for generalizing the multi-objective tness assignment, the Strength Pareto approach.
Responsible Party
Creator(s):Author: Tanja Zerenner
Publisher:CRC/TR32 Database (TR32DB)
TR32 Topic:Atmosphere
Subject(s):CRC/TR32 Keywords: PhD Report
File Details
File Name:Report3_Zerenner_2013.pdf
Data Type:Text
File Size:1933 kB (1.888 MB)
Date(s):Available: 2013-08-28
Mime Type:application/pdf
Data Format:PDF
Download Permission:OnlyTR32
General Access and Use Conditions:For internal use only.
Access Limitations:For internal use only.
Licence:TR32DB Data policy agreement
North:-no map data
Measurement Region:Other
Measurement Location:--Other--
Specific Informations - Report
Report Date:28th of August, 2013
Report Type:PhD Report
Report City:Bonn, Germany
Report Institution:Meteorological Institute, University Bonn
Number Of Pages:15
Period of Pages:1 - 15
Further Informations:TR32 Student Report Phase II
Metadata Details
Metadata Creator:Tanja Zerenner
Metadata Created:2013-12-04
Metadata Last Updated:2013-12-04
Funding Phase:2
Metadata Language:English
Metadata Version:V40
Dataset Metrics
Page Visits:291
Metadata Downloads:0
Dataset Downloads:1
Dataset Activity