Research Projects

Heavy Precipitation over complex terrain

Local orography governs the spatial distribution of precipitation over complex terrain both for annual sums as well as for single events. Intense, lengthy precipitation events are typically found on the upwind side of the mountains, while shorter events with less intensity are characteristic for the lee side. Besides the orography the rain enhancement is governed also by several atmospheric parameters like horizontal wind velocity and stability of the stratification. Therefore the frequency of regional flooding, which is primarily a consequence of intense stratiform precipitation with long duration, differs from region to region and from situation to situation.
The detailed spatial distribution of particularly heavy rainfall has important applications and consequences for hydrology, water management and agriculture. Since most of the existing rainfall observation networks have a low density (rain gauge) or a short observation period (radar), additional model simulations are necessary to derive climatologic rainfall fields with a high spatial resolution. The aim of this project is to interpolate measured rainfall data from coarser rain gauge networks to a finer mesh with simulations of a numerical model. With this, it is possible to derive high resolved climatological rainfall statistics.

Issues and Goals:

The goal of the project is twofold: The first objective is to analyse and identify the atmospheric conditions that favour events of heavy precipitation (stratiform and convective) over complex terrain. This includes also the study of dynamical effects such as flow splitting, channelling or the flow over and around an obstacle, all in situations during rainfall.
The second and main point of the work is to assess the spatial distribution of (stratiform) heavy rainfall with long durations climatologically. Like mentioned above, this could be done in detail only with simulations from a numerical model. Since the purpose is to simulate lots of single events, the model for the orographically induced rainfall needs to have short computing time, simple use and easy initialisation from vertical sounding data.
With this model a regionalization of single rain gauge measurements will be performed. The simulation of past events over a long period leads to an estimation of the probability of extreme events and the potential of hazards with high spatial resolution.
The areas under investigation are the low mountain ranges of the Black Forest, the Palatinate Forest and the Swabian Alb in Southern Germany. However, it is generally possible to transfer the methods and results to other low mountain ranges.

Methods:

The atmospheric conditions, which govern heavy precipitation events, both stratiform and convective, are analysed statistically using correlations analysis techniques and frequency distributions from a multiplicity of parameters derived from radiosounding data.
For the investigation of rainfall characteristics concerning the spatial distribution (amount of the orographically induced rainfall, wind-/leeward effects, wind drift etc.), observation data from several networks (DWD stations, LFU/UMEG ombrometer data, radar data) are analysed.
To assess the spatial distribution of stratiform heavy rainfall in Southwest-Germany with respect to its duration, a numerical model is developed for the interpolation of rainfall from coarser rain gauge network. This physical model is based on the linearized, steady und frictionless equations of motion assuming hydrostatic balance to give the 3-D topographically forced vertical motions with respect to the Froude-Numbers (dimensionless ratio between wind velocity and stability of the stratification and a scale from terrain) and the wind direction. The computed upward motions causes specific condensation rates and total rainfall rates using temperature and humidity profiles from radiosounding data. Effects such as wind drift of the hydrometeors and lee drying are incorporated in the model. Due to the short computing time and the simple initial and boundary conditions it is possible to simulate single events of heavy rainfall over a long past period.

Results:

Heavy rainfall is frequently associated with wet and warm air mass in the lower layers (up to 850 hPa). In the layers above (olsi 500 hPa), there couldn't be found a significant indicator in most cases. Analyses also show a wind direction between southwest and northwest on the synoptic scale. Parameters describing the stability of the boundary layer (Brunt-Väisälä-Frequency) or the buoyant energy (CAPE, Rib) show a distinct difference between days with convective heavy precipitation and days with moderate precipitation.
Generally, the horizontal distribution of stratiform heavy rainfall with long duration is strongly determined by orographic parameters, mainly elevation and slope. Therefore, intense stratiform precipitation events with long duration are much more frequent on the windward side of Black Forest. Due to the lee effect and the fewer slopes, stratiform heavy rainfall is seldom in the region of the Swabian Mountains. In contrast, convective heavy rainfall could be observed over the whole region, somewhat more frequently at the top of Black Forest Mountains. Radar data show that after the formation, the lifetime of a convective cell and therewith the precipitation field is a stochastic process, depending on the horizontal wind direction, but not on the orography.
The advection of humidity as well as the dimensionless Froude-Numbers control the orographic enhancement of precipitation. The strongest enhancement is found at situations with higher Froude-Numbers, meaning that most of the airflow will go rather over than around an obstacle. This finding is obvious on the average over several stratiform events as well as during single events.
In spite of the simple features of the orographic rainfall model (e.g. no explicit microphysical parameterization), first simulations seem to be quite realistic. Rainfall distributions are qualitatively in good agreement with observations. It becomes evident, that the consideration of effects such as wind drift due to formation and fall time of the hydrometeors and lee drying increase the skill of the model substantially. Even though the first simulations are quite realistic, extensive evaluation of the model and adjustment of the parameters are still necessary.

Person
Michael Kunz

Advisor
Prof. Kottmeier

Publikationen

Short Text