Ezio Todini 70th Symposium: my talk

Please you will find below the picture (click on it!) the presentation I will give the 6th of June in Bologna to mark the 70th birthday of Prof. Ezio Todini. I will say things that I already said in different contexts. However, reading them in this form can help to understand them more.

Ezio did a lot of interesting works on rainfall runoff modelling, starting from the Arno model, that can be considered a milestone in the field and Topkapi, his distributed hydrological model. He contributed also to data assimilation, models calibration, time series analysis, and I will dedicate a post in a few day to him and to his most important papers.

Below you can find also the link to our posters about:

• The JGrass-NewAGE system
• The GEOtop model embedded in OMS3

The topics of the first poster where already covered in various posts. The second poster covers the same topics of the paper for the Hydrology days conference.

We need water

This time just a photo by the master Sebastiao Salgado. A photo that says more than thousand words.

The Snow Water Equivalent (NewAGE-SWE) model component in JGrass-NewAGE

This is, at the moment, a stub prepared to contain the material related to the paper provisionally entitled: The Cache La Poudre snow water equivalent modelling with JGrass-NewAGE.

The paper has been finally submitted to GMD, and the Discussion Manuscript is therefore visible to everybody on GMDD. The revised version of the paper can be seen instead here.

Here the abstract of the revised paper: "The paper presents a package of a modified temperature index based snow water equivalent model as part of the hydrological modeling system NewAge-JGrass. 35 Three temperature-based snow models are integrated in the
NewAge-JGrass modeling system and use many of its com- ponents such as those for radiation balance (SWRB), kriging (KRIGING), automatic calibration algorithms (particle swarm optimization), and tests of goodness of fit (NewAge-V), to build suitable modelling solutions (MS). Similarly to all the NewAge-JGrass components, the models can be exe- cuted both in raster and in vector mode. The simulation time step can be daily, hourly or sub-hourly, depending on user needs and availability of input data. The MS are applied on the Cache la Poudre river basin (CO, USA) using three test applications. First, daily snow water equivalent is simulated for three different measurement stations for two snow model formulations. Second, hourly snow water equivalent is sim- ulated using all the three different snow model formulations. Finally a raster mode application is performed to compute snow water equivalent maps for the whole Cache la Poudre basin. In all the applications the model performance is satis- factory in terms of goodness of fit relative to measured snow water equivalent time series and the results, and the differences in performances of the different modelling solutions are discussed."

The model, the data, and the simulations' scripts used in the paper will be made available through the appropriate links from this page in order to let people reproduce our results. In fact, we call it Reproducible Research. It is certainly a challenging strategy for any researcher, and sometimes data cannot be disclosed. However, I said what I said, and this is a couple of old posts which fully explain my position (About Scientific Software, No code - No paper).

A summer school on landslides modelling (July 4-10 2013)

Shallow landslides are a widespread hazard on which I did a few selected publications. Modelling landslide occurence is problem that is difficult to grasp also because knowledge about the phenomenon covers geology, geotechnics-geo-mechanics , agronomy (especially regarding the role of roots) and, obviously, hydrology. As a result a modern landslide expert has to get accustomed with several point of views which are often divergent, and various modelling approaches and numerics, which makes uneasy a synthesis. Historically the field has crystallised along some concepts and methods which are summarised in one of my previous posts (Guidelines for the mapping of the triggering of landslides and debris flow). To investigate new ideas, mix the competences, and summarise good old tools, University of Calabria and CUDAM (of University of Trento) organised a summer school of which you can find information in the link below the picture.

Lectures of the school were:

Dino Bellugi, Matteo Berti, Giambattista Chirico, Ning Lu, Riccardo Rigon, Cristina Rulli.

The twenty participants came from nine nations and were selected from doctoral students and post-docs.
In the summer school website, it is here, you can find the lectures and the videos of the school.

Using ggplot2 for Hydrology

Leonardo Perathoner prepared this for helping my bachelor students:

"Since some student asked to me some information about ggplot, I write it here for everybody.

If you are interested to plots and a graphic quality superior to the standard R, you can utilize the R package ggplot2 !!

Here they are some resources that I want to I would advise to use in your learning process:

1) First a good manual

http://www.bioinformaticslaboratory.nl/twikidata/pub/Education/ComputinginR/ggplot2-book.pdf 

and then, these other sites about ggplot I often use:

2) the official site:

http://docs.ggplot2.org/current/ 

3) a quick cookbook with examples and code snippets

http://www.cookbook-r.com/Graphs/ 

Leonardo "

PS: many other resources here:

http://abouthydrology.blogspot.it/2012/08/r-resources-for-hydrologists.html
http://abouthydrology.blogspot.it/2016/09/cool-r-tydyverse.html

A Geomorphological Assessment of Constable Landscapes

I took this post from the River Management Blog, and the post is by Simon Dixon. John Constable is one the greatest English landscape painter, and, in the words of the blog authors, contributed very much to the idea of what a natural landscape is. However, reading the post, you can realize that his paintings portrait not so natural landscapes.

Click, as usual on the painting to jump on the post.

What is the minimal geomorphology-based hydrological model?

This is one of the objects of the research Alban de Lavenne, a Ph.D. student of Christophe Cuddenec pursued here in Trento, during his three months stay the last Fall. Similar approaches can be found in the work by Fenicia, and especially in Fenicia et al., 2008. Fenicia 2008, is a must-read paper, since it is well written, and smart. However there, the  Authors did not use any geomorphic information as guideline in their modeling but just a scheme based on reservoirs, which is, in my view, out-of-date. Which is kind of a pity, considering that we know very much of the morphology of river, even when they are ungauged. Thanks to SRTM, ASTER, and other topographic data, the topology and geometry of river networks, and Earth's elevation is known with unsurpassed precision all over the World, and therefore, the first step to predict what happens in ungauged basins would be to use their geo-morphology, which, in fact, was shown to be important in many papers.

Alban used in his investigation a few simple geomorphological unit hydrograph schemes based on the width function:

• 1F1U1P: 1 velocity applied to entire flow path length
• 1F2U2P: 2 velocities (respectively on hillslope and channelized length) -
• 2F2U3P: 1 velocity for surface flow, 1 velocity for subsurface flow
• 2F3U4P: 2 velocities for surface flow, 1 velocity for subsurface flow

In this context, "velocity" can be read also as "mean travel time" since the relation between the two quantities is given by the appropriate lengths measured along the flow paths.

In any case, the poster you can retrieve below the figure (click on the figure) tells it all.