Chwala, C., Gmeiner, A., Qiu, W., Hipp, S., Nienaber, D., Siart, U., Eibert, T., Pohl, M., Seltmann, J., Fritz, J., and Kunstmann, H.: Precipitation observation using microwave backhaul links in the alpine and pre-alpine region of Southern Germany. In: Hydrol. Earth Syst. Sci., 16, 2647-2661, doi:10.5194/hess-16-2647-2012, 2012
Winter School: Microwaves Links and Precipitation,
February, 27th - March, 2nd, 2012, Garmisch-Partenkirchen (FLYER)
Virtual Institute PROCEMA
(Regional Precipitation Observation by Cellular Network Microwave Attenuation and Application to Water Resources Management)
Estimating the spatial and temporal distribution of precipitation is of crucial importance for hydrological analyses. This is particularly true in regions with high spatial precipitation variability. A new means to accomplish this task is exploiting attenuation data from commercial backhaul links, which is a useful complement to traditional rain gauge and radar derived estimations, since it is based on a different spatial and temporal scale.
- apply this new method in an alpine region
- improve its results
- merge data with both traditional methods (radar, gauge)
- and further improve the understanding of its physical effects.
The project work is divided into the following four packages
Our test area is the alpine region around Garmisch-Partenkirchen in southern Germany. The terrain is characterized by complex orography which makes accurate radar and gauge measurements difficult. We installed GSM data logger modules at five microwave backhaul link towers in the test area. They record minutely averages from an analog voltage output of the link receiver unit. Data is then transferred via GSM to a data base for further processing.
The attenuation along those microwave links is not only caused by precipitation but also by other atmospheric parameters e.g. humidity and temperature, on a smaller level tough. We thus developed an algorithm to detect the attenuation periods caused by rain. After that classification we are able to calculate rain rates from the link’s attenuation data. The figure on top of this page shows a comparison of data obtained from the link leading from Garmisch to Mount Laber near Oberammergau with the two nearby DWD rain gauges.
Map of the test area around Garmisch-Partenkirchen showing the location of the gauges and microwave backhaul links.
Precipitation estimates derived by microwave attenuation of commercial link
Our new approach to measure rainfall using the attenuation data from terrestrial microwave links provides some advantages over the established method with radar and rain gauge. But still radar and rain gauges have their strengths too. Thus a combination of all the three focusing for each on its advantages provides the most accurate information for rainfall.
We develop a statistical method using Copula Theory to combine the three data sources into one spatial rainfall field. The advantage of Copulas is that they do not only give a measure of the correlation of two data sets, but also yield a full probability density function (pdf) for each correlated data point.
To study the interaction of hydrometeors (rain, snow, sleet, hail) and other parameters of the atmosphere (temperature, water vapor) with microwave radiation in more detail an own transmission experiment was designed and built. It is operated at the TERENO test site in Fendt near Weilheim.
The system operates at two frequencies, 22 GHz and 35 GHz, with altering horizontal and vertical polarization. This enables differential measurements in frequency and in polarization which reduces instrumental influences and yields information about the drop size distribution. For the systems configuration a monostatic setup with a large trihedral reflector was chosen to make it fully coherent. It thus provides stable phase information which gives a very good measure for the line integrated absolute humidity along the transmission path. In addition the system with its high sampling rate also allows a first study of the rain induced signal fluctuations which are supposed to provide information about the precipitation type and the drop size distribution
The transmitter/receiver unit and the laser distrometer on the 10m tower
In addition to the experimental study a simulation tool was developed to study the fluctuation processes induced by precipitation.
Over 6 million falling drops are simulated for a synthetic transmission setup with 22 GHz and a path length of 500m. Drop sizes are chosen according to the Marshal-Palmer distribution. The simulated incoming plain wave is scattered by all drops following analytic Mie scattering formulas. At the receiving end the resulting signal noise is recorded in phase and amplitude, including polarization information. First results show a significant influence of the rain rate on the Doppler spectrum of the received signal.
Schematic of the simulated transmission setup with rain drop populating the volume
Prof. Dr. Pinhas Alpert and Prof. Dr. Hagit Messer, Tel Aviv University (TAU), Department of Geophysics and Planetary Sciences and School of Electrical Engineering, Tel-Aviv 69978, Israel
Email: pinhas, ∂post tau ac ilmesser∂eng tau ac il
Dr. Alon Rimmer, Israel Oceanographic & Limnological Research Ltd, Kinneret Limnological Laboratory (KLL), P.O.B. 447, Migdal 14950, Israel
Email: alon∂ocean org il
Dr. Jörg Seltmann, German Weather Service (DWD), Meteorological Observatory Hohenpeissenberg MOHp, 82383 Hohenpeissenberg
Email: joerg seltmann∂dwd de
Prof. Dr. Martin Pohl, University of Applied Science Regensburg (UAS-R), Section of Applied Mathematics, Universitätsstraße 31, 93053 Regensburg
Email: martin pohl∂mathematik fh-regensburg de
Prof. Dr. Manfred Lange, The Cyprus Institute, Energy, Environment and Water Research Center (EEWRC), P.O. Box 27456, CY-1645 Nicosia, Cyprus
Email: m a lange∂cyi ac cy
PROCEMA is funded by the HGF Impuls- und Vernetzungsfond (VH-VI-314)