A Gaussian Markov random eld approach for radar rainfall information

This page lists all metadata that was entered for this dataset. Only registered users of the TR32DB may download this file.

Feature
Request downloadRequest download
Full Name:
Affiliation:
eMail:
Purpose of use:
 
I accept the Data protection statement.
Bot check:
Type all characters with this
color
.
 
It is case sensitive.
 
 
↻
 
Submit
Citation
Citation Options
Identification
Title:Main Title: A Gaussian Markov random eld approach for radar rainfall information
Description:Abstract: Spatially distributed, high-resolution precipitation rates are key ingredients for modeling soil-vegetation processes, water and solute transports in mesoscale catchments, and for short-range weather prediction. The ultimate goal of our study is to develop a space-time, multilevel statistical model that merges rain radar measurements with other observations of precipitation. This is a challenging task since it aims at combining data sources with a variety of error structures, and temporal resolutions. E.g., in-situ measurements are quite accurate, but available only at sparse and irregularly distributed locations, whereas remote measurements cover complete areas but su er from spatially and temporally inhomogeneous systematic errors. The rst step towards such a space-time precipitation model is to develop a statistical model for precipitation based on radar measurements. Precipitation rates over a region of about 230 x 230 km2 are provided by a composite of the two polarimetric X-band radars in Germany. The two radars are located in a distance of about 60 km in Bonn and Jülich, respectively. For the statistical model formulation we use a Gaussian Markov random eld as underlying process. A Markov random eld is a suitable model to account for spatial dependencies if the region where dependencies are observed can be reduced to a small neighborhood without losing information. This makes large data problems computationally feasible, since the neighborhood structure stands in one-to-one relation with a sparse precision matrix. We start with the spatial analysis of rainfall intensities derived from radar reectivities. The images consist of 460 x 461 points with a resolution of 500m x 500m that lie in the overlap of the Jülich and Bonn X-band radars and cover the Rur catchment. We derive a stationary and isotropic GMRF by tting its correlation function to the empirical correlation function of the data.
Responsible Party
Creator:Katharina Krebsbach (Author)
Publisher:CRC/TR32 Database (TR32DB)
Publication Year:2013
Topic
TR32 Topic:Atmosphere
Related Subproject:D5
Subject:Keyword: PhD Report
File Details
Filename:Report3_Krebsbach_2012.pdf
Data Type:Text - Text
File Size:324 KB
Date:Available: 30.09.2012
Mime Type:application/pdf
Data Format:PDF
Language:English
Status:Completed
Constraints
Download Permission:Only Project Members
General Access and Use Conditions:According to the TR32DB data policy agreement.
Access Limitations:According to the TR32DB data policy agreement.
Licence:[TR32DB] Data policy agreement
Geographic
Specific Information - Report
Report Date:30th of September, 2012
Report Type:PhD Report
Report City:Bonn, Germany
Report Institution:Meteorological Institute, University of Bonn, Germany
Number of Pages:7 (1 - 7)
Further Information:TR32 Student Report Phase II
Metadata Details
Metadata Creator:Katharina Krebsbach
Metadata Created:16.12.2013
Metadata Last Updated:16.12.2013
Subproject:D5
Funding Phase:2
Metadata Language:English
Metadata Version:V50
Metadata Export
Metadata Schema:
Dataset Statistics
Page Visits:757
Metadata Downloads:0
Dataset Downloads:5
Dataset Activity
Feature
A download is not possibleDownload
top ∧