ERC grants in Water Related topics

To my knowledge, three were, so far the grants given in water related topics by the ERC committes.

The first was assigned to Andrea Rinaldo's in 2008 and was entitled on "Modelling waterborne epidemics. It was the first ERC grant to be assigned to a hydrologist, and the success story can be found at the link above. His research was also told at the opening of our 2014 Doctoral School.

The second one was given to Gunther Bloeschl in 2012, and is entitled: Decipher River Flood Change. Its short description is:
" Major floods around the world have raised questions about the frequency and magnitude of such phenomena. Although changes in climate and land use are known to play a critical role in river floods, how they actually translate into considerable variations in intensity remains unknown "

The third ERC was assigned in 2013 to Doerthe Tetzlaff for a project called veWA:

"VeWa: Vegetation effects on water flow in high-latitude ecosystems” project will examine the impacts of climate change on vegetation-water linkages along a northern climatic gradient, investigating four intensively studied experimental sites in the UK, Canada and Sweden ...  Such a geographically extensive comparison has never been conducted in such environments and will allow the consistency of processes, drivers and climatic impacts to be assessed across a range of spatial scales”

Installing MeteoIO on Linux

Matteo Dall'Amico (see previous post on GEOtop installation) also provided instructions for compiling MeteoIO on Linux). For Mac OS instructions, please see this post.

# Check the environment compiler
sudo apt-get install make
sudo apt-get install g++


# Install the following packages (or verify you have them already). For Linux:


# install SVN
sudo apt-get install subversion

# install cmake
sudo apt-get install cmake
sudo apt-get install cmake-curses-gui

# install gdal libraries
sudo apt-get install libgdal-dev

# install proj libraries
sudo apt-get install proj-bin
sudo apt-get install libproj-dev

# After having selected your directory for installing meteoIO libraries with the usual console commands
# download the source code from the SVN:

svn co https://models.slf.ch/svn/meteoio/trunk

# then go inside the trunk folder

cd /path_to_MeteoIO_SVN/trunk/

# and type the cmake command:

ccmake .

# Press enter in the console and then "c" to configure:
Please set to ON:

BUILD_SHARED_LIBS
PLUGIN_A3DIO
PLUGIN_ARCIO
PLUGING_ARPSIO
PLUGIN_GEOTOPIO
PLUGIN_GRASSIO
PLUGIN_PGMIO
PLUGIN_SMETIO
PLUGIN_SNIO
PROJ4 (then you have to verify the paths to LIBPROJ and to LIBPROJ_INCLUDE_DIR)

All the rest can be set to OFF.

# Then, press "g" to generate the makefile and exit from the console.
# In the terminal, type:

make install

# (with "sudo" if you want to install it as super user).
# Eventually, you have to make the system aware of the new library by typing:

sudo ldconfig -v

After having installed meteoIO, in the needs to understand what it does, one can browse the examples in the "doc" directory. However, s/he can also built the doxygen documentation. Please be sure that you have Doxygen installed. For doing it, please first check you have doxygen installed. Subsequently:

• using a terminal move to the directory where the meteoIo source code is located. 
• execute "doxygen -g <configfile>" to create a configuration file, "configfile" for the project
• execute doxygen <configfile> to create the documentation
• Find the documentation in a new html directory created
• Access it by clicking on the index.html file

Installing GEOtop 2.0 on Linux

I receive from Matteo Dall'Amico and I publish the following instruction to compile GEOtop 2.0 from command line on Linux.

# Install the following packages (or verify you have them already):

# install GIT
sudo apt-get install git

#Install and compile MeteoIO libraries
See instructions here

#Install pkg-config
sudo apt-get install pkg-config

# install boost libraries
sudo apt-get install libboost-dev
sudo apt-get install libboost-test-dev
sudo apt-get install libboost-regex-dev
sudo apt-get install libboost-program-options-dev
sudo apt-get install libboost-filesystem-dev
sudo apt-get install libboost-iostreams-dev

# install autoconf
sudo apt-get install autoconf

# download GEOtop
git clone https://github.com/skyglobe/geotop#

# go inside the geotop folder
cd geotop

# create the makefile
ccmake .

# add the path of MeteoIO: /usr/local
# To process the file press "c" (like "configure") and then "g" (like "generate")
# make the project
make

#try geotop example
cd tests/test_sample_run/small_example

#create symbolic link
ln -s ../../../geotop/geotop geotop_TN

#run geotop_TN
./geotop_TN .

Richards equation (and hillslope hydrology) video from the Summer School on Landslides

Last summer school on landslides that University of Calabria Organised was pretty successful.  Besides organising it together with Giovanna Capparelli and Giuseppe Formetta, I also gave a lecture on Richards equation (you can find the slides here). Finally, thanks to the organisation, the videos of my lecture are available.

The first part

                                       

The second part

The third part

WRF and LES

Dino Zardi sent to me and some other colleagues, a small synthesis WRF and LES which you can find below

Introduction to Large-Eddy Simulations (LES) with the Weather Research and Forecasting (WRF) model,

By Lorenzo Giovannini

The Weather Research and Forecasting model (WRF, Skamarock et al., 2008) is a numerical weather prediction model used for both operational and research applications at scales ranging from regional to large-eddy simulations (LES). Several planetary boundary layer (PBL) parameterization choices are available in WRF, which are intended for use in cases in which the horizontal grid resolution does not enable the representation of three-dimensional turbulence (resolution coarser than 1 km). For simulations requiring higher resolution to capture evolving three-dimensional turbulence or flows in complex terrain, a LES is more appropriate. LESs explicitly resolve large turbulent eddies, whereassome portion of the turbulence (below the filter imposed by the computational grid) must still be modeled using a sub-grid scale (SGS) model. There are different options in WRF to model SGS turbulence. In the first option eddy viscosity coefficients are determined using a 3D Smagorinsky turbulence closure from deformation and stability, whereas in the second option a prognostic equation for turbulent kinetic energy (TKE) is used, and eddy viscosity coefficients are based on TKE. Finally, another approach to model SGS stresses in WRF is the nonlinear backscatter and anisotropy (NBA) model of Kosović (1997), (see Skamarock et al., 2008 and Kirkil et al. 2012 for further details). 

WRF (in LES mode) calculates horizontal and vertical diffusion in physical space.

Several issues arise when performing LES simulations, and in particular LES simulations nested in mesoscale domains, for which the flow contains essentially no resolved turbulence. To this regard several authors provided recommendations that should be followed. As discussed in Mirocha et al. (2010), the aspect ratio, or ratio of the horizontal grid size to the vertical grid size, can strongly impact the accuracy of an LES. They found that aspect ratios between 2 and 4 are ideal for neutral boundary layer flows over a flat plate, but the authors pointed out that different results may be expected in situations with more complex terrain. Then it must be remarked that the grid spacing used in the LES domain must be << than the energy-containing eddies, to resolve large eddies. This recommendation may be a constraint in stable atmospheric boundary layer (ABL), when the expected large eddies are of the order of 10 m. Finally, one of the most important issues arising when coupling an LES with a mesoscale model is the spin-up of turbulence at inflow boundaries, as it is discussed also below.

Moeng et al. (2007) pointed out that WRF can be a reliable tool to perform real-world LES. In fact WRF uses real world terrain and land use data as well as real atmospheric conditions. In order to test the performance of the two-way nesting capability for LES in WRF, the authors performed LES-within LES simulations, where one LES was nested within the other. Promising results were obtained with the alteration of the SGS stress model, the careful selection of the nest size, and the use of a relaxation zone close to nest edges. A similar investigation was carried out by Mirocha et al. (2013), who in particular analyzed errors arising at nest interfaces by changes in mesh spacing from a coarser domain to a finer domain, using different SGS models. They found significant discrepancies in many parameters between the nested simulation and a single-domain simulation, concluding that extensive buffer zones, whose extension depends on the SGS model used, are required for equilibration of flow parameters on nested domains. Moreover the authors stated that vertical nesting, which is not present inthe current WRF release, would likely improve the results. Both Moeng et al. (2007) and Mirocha et al. (2013) pointed out that nesting an LES inside a mesoscale model would pose a much bigger challenge than nesting a finer LES within a coarser LES, due to the different ways to treat turbulent motions and due to turbulence spin-up problems when prescribing an inflow into the LES domain from the mesoscale model, whose flow is laminar by construction.

This problem was analyzed in detail by Mirocha et al. (2014), who highlighted that, despite mechanisms for downscaling from mesoscale to LES are present in several models (including WRF), little information exists regarding conditions under which this approach is appropriate. In particular the authors compared the development of turbulence between LES nested within mesoscale domains and non-nested LES. They found that, under relatively weak forcing conditions, turbulence generally develops too slowly in LES nested within mesoscale domains. Moreover they found that nesting a fine LES domain within a coarse LES gave better results than nesting it directly within the mesoscale model, as turbulence can begin to develop in the coarser LES.

Applications of nested LES within mesoscale domains using WRF in real test cases are provided byLiu et al. (2011) and Talbot et al. (2012). Liu et al. (2011) tested WRF-LES in a nested mesoscale-LES configuration for wind power applications down to a resolution of ~100 m. Four coarse domains were run with mesoscale model settings and continuous data assimilation, while two finer mesh domains were run with LES model settings. The model results, compared with wind farm anemometer measurements, were found to capture many intra-farm wind features and microscale flows. Talbot et al. (2013) tested the skills of WRF-LES in nested real-world simulations using six nested domains, with a resolution of 50 m in the inner domain. The three largest domains were run with mesoscale model settings, while the three inner domains were run with LES model settings. The performance of the model was rather poor in simulating wind speed and direction, while better results were found with respect to air temperature. Several sensitivity tests highlighted that the most important controls on model results were the meteorological forcing data, which are downscaled by WRF and significantly impact also the smaller domains. In particular the errors in wind speed and direction in the mesoscale simulations were largely passed on to all LESs, which were unable to correct them. On the other hand it was found that the mesoscale horizontal turbulence closure and the LES SGS model had a relatively small impact on results. Finally increased resolution improved the ability of the model to capture surface variability, but did not improve regionally averaged results or bulk atmospheric boundary layer properties.

Kirkil G, Mirocha J, Bou-Zeid E, Chow FK, Kosović B. 2012. Implementation and evaluation of dynamic subfilter-scale stress models for large-eddy simulation using WRF. Monthly Weather Review140: 266-284.

Kosović B. 1997. Subgrid-scale modeling for the large-eddy simulation of high-Reynolds-number boundary layers. J. Fluid Mech. 336: 151-182.

Liu Y, Warner T, Liu Y, Vincent C, Wu W, Mahoney B, Swerdlin S, Parks K, Boehnert J. 2011. Simultaneous nested modeling from the synoptic scale to the LES scale for wind energy applications.Journal of Wind Engineering and Industrial Aerodynamics 99: 308-319.

Mirocha JD, Lundquist K, Kosović B. 2010. Implementation of a non-linear subfilter turbulence stress model for large-eddy simulation in the advanced research WRF model. Monthly Weather Review 138:4212-4228.

Mirocha JD, Kirkil G, Bou-Zeid E, Chow FK, Kosović B. 2013. Transition and equilibrium of neutral atmospheric boundary layer flow in one-way nested large-eddy simulations using the Weather Research and Forecasting model, Monthly Weather Review 141: 918-940.

Mirocha JD, Kosović B, Kirkil G. 2014. Resolved turbulence characteristics in large-eddy simulations nested within mesoscale simulations using the Weather Research and Forecasting model. Monthly Weather Review 142: 806-831.

Moeng C-H, Dudhia J, Klemp J, Sullivan P. 2007. Examining two-way grid nesting for large eddy simulation of the PBL using the WRF model. Monthly Weather Review 135: 2295-2311.

Skamarock WC, Klemp JB, Dudhia J, Gill DO, Barker DM, Duda MG, Huang X-Y, Wang W, Powers JG. 2008. A description of the advanced research WRF version 3. NCAR Technical Note TN-475+STR, 125.

Talbot C, Bou-Zeid E, Smith J. 2012. Nested mesoscale large-eddy simulations with WRF: performance in real test cases. Journal of Hydrometeorology 13: 1421-1441.