All posts by Javier Garcia Hernandez

Courses about ‘Hydrological and hydraulic modelling with RS MINERVE’ in Lima (Peru) – April 2015

Two courses (beginner and advanced) about ‘Hydrological and hydraulic modelling with RS MINERVE’ have been held on April 13th to 15th and on April 16th and 17th 2015 in the Headquarters of the National Service of Meteorology and Hydrology of Peru (SENAMHI), in Lima, Peru.

In the first course, for beginners, theoretical and practical parts have been combined to provide the required knowledge to model hydrological basins with the help of RS MINERVE. 35 participants from different Peruvian organizations and consulting enterprises have taken part with a great interest and success.

In the second course, for advanced users which already followed the course for beginners in 2014 or 2015, the details of the RS GIS and the RS Expert modules where revealed. Then, a practical exercise allowed the participants to become experts in hydrological and hydraulic modelling with RS MINERVE. About 25 participants took part in this course, with a big success that predicts another time two courses for 2016.

The courses, taught by J. García Hernández and J. Fluixá Sanmartín, from the Centre de recherche sur l’environnement alpin (CREALP-Switzerland) and by J. Paredes Arquiola, from Technical University of Valence (UPV-Spain), was organized in collaboration with the SENAMHI, the CREALP, the University of Zurich, the Swiss Agency for Development and Cooperation (SDC), Meteodat, Care Perú, ProyectoGlaciares and the Ecole Polytechnique Fédédale de Lausanne (EPFL).

More information (in Spanish) on this webpage.

 

New algorithms for hydrological parameters optimization

In addition to the SCE-UA (Shuffled Complex Evolution – University of Arizona), used for the optimization of the hydrological parameters in RS MINERVE, two new algorithms have been implemented: the “Uniform Adaptive Monte Carlo” and the “Coupled Latin Hypercube and Rosenbrock”

The Uniform Adaptive Monte Carlo (UAMC) algorithm is based on the Monte Carlo experiments that rely on repeated random sampling but has been modified in order to iteratively adjust the solution space.
The algorithm randomly launches a collection of simulations (block) and finds the better results in the solution space. Afterwards, the solution space is adjusted and a new group of simulations starts. The process is repeated until the optimization converges to the best set of parameters.

The Coupled Latin Hypercube and the Rosenbrock algorithm, generates a powerful tool for optimization of complex problems. This combined algorithm can discretize a wide domain and then narrow your search to smaller sectors. Scanning of the space of possible solutions is performed by the Latin Hypercube. This algorithm allows pseudo-statistical sampling conditioned by the previous calculated solutions. The Latin hypercube is an evolution of the Monte Carlo method, with more homogeneous samples achieved with fewer samples. An important advantage of this method is that the dimension of the problem is defined by the division of the latin hypercube and not by the number of parameters.
The best results from samples become the starting points required for Rosenbrock algorithm. The advantage of this subroutine calculation lies in the speed to obtain near optimal values. This algorithm is based on a gradient search, adjusting axis changes based on the direction of maximum enhancement, thus reducing the number of evaluations of the objective function.

More details in the technical manual.

 

Course ‘Hydrological and hydraulic modelling with RS MINERVE’ in Sion (Switzerland) – November 2014

The third installment of the course ‘Hydrological and hydraulic modelling with RS MINERVE’ in 2014, after those of Lima (Peru) and Parana (Argentina), was held on November 19th and 20th in the “Casernes” of Sion, Switzerland. Following a now proven pattern, theoretical and practical sessions were combined to provide the knowledge required to model hydrological basins with the help of RS MINERVE. As a direct example of application, the MINERVE operational system for flood forecasting and management in the Upper Rhone River basin was also presented. 15 participants, mainly civil servants associated with the Canton of Valais, took part with great interest and success. .

The course, taught by Javier García Hernández, Alain Foehn, Samuel Alesina and Javier Fluixá Sanmartín, from the Centre de recherche sur l’environnement alpin (CREALP) and some members of the Steering committee of Hydro10, was organized by the CREALP in collaboration with the Ecole Polytechnique Fédérale de Lausanne (EPFL), the Technical University of Valencia (Spain) and the Hydro10 Association.

Vectorial icons for hydrological-hydraulic objects

Hydro10 collaborates in the development of RS MINERVE and has created a new graphic line with vectorial images for all available objects  in RS MINERVE.

The attractiveness of a software comes not only from a robust mathematical basis but also from an enjoyable interface. The new icons will improve the quality of the user-friendly graphical interface and the details when zooming. This new development reinforces the look of the software and enhances the significant work achieved about new modules  in recent months.

 

Stochastic simulation module

Hydro10 collaborates in the development of RS MINERVE to make the software accessible to a larger number of users. In this context, Hydro10 has developed a new module, called “Stochastic simulation”. This new module is capable of generating a set of simulations based on different parameters or initial conditions with values located in a random interval defined by the user.

This module provides, for each element of the model, simulated hydrographs, corresponding statistics (mean, median, quartiles, minimum and maximum values), and related set of parameters for each simulation.

The new version of RS MINERVE, v1.2.1, already includes this new module that you can test from now.

 

Potential Evapotranspiration calculation

RS MINERVE currently offers the possibility to calculate the Potential Evapotranspiration (ETP) values from different methods when no ETP data is available.

The 4 implemented ETP methods are :

  • Turc: calculation of ETP depending on the global radiation Rg (based only on latitude and longitude) and temperature
  • McGuinness: method depending on Rg (based only on latitude and longitude) and temperature
  • Oudin: a french method depending on the extra-terrestrial radiation Re (based only on latitude) and T
  • A uniform ETP

The value of global radiation Rg is a monthly averaged value which depends on the latitude and longitude of the hydrological model.   These Rg data were obtained from the Surface meteorological and Solar Energy (SSE) web portal, sponsored by the NASA’s Applied Science Program (http://eosweb.larc.nasa.gov/sse).

Rg data takes into account 22 year monthly average (July 1983 – June 2005). The latitude and the longitude values indicate the lower left corner of a 1×1 degree region. Negative values are south and west; positive values are north and east. Boundaries of the -90/-180 region are -90 to -89 (south) and -180 to -179 (west). The last region, 89/180, is bounded by 89 to 90 (north) and 179 to 180 (east). The mid-point of the region is +0.5 added to the latitude/longitude value. These data are regional averages, not point data.

The value of extra-terrestrial radiation Re is directly computed based on the latitude of the hydrological model and parameters such as sun-earth distance and solar declination.

For more information, please refer to the RS MINERVE Technical Manual.

 

Time-slice simulation and scenario simulation

Two new modules, specifically created for research institutes or hydrologic/hydraulic projects, have been added to the software RS MINERVE and allow to easily evaluating hydrologic and hydraulic results.

Time-slice simulation facilitates the analysis of large data sets without overloading the computer memory.

Scenario simulation introduces the possibility of simulating multiple weather scenarios or several sets of parameters and initial conditions to study the variability and sensitivity of the model.

Module for automatic calibration of hydrological parameters

A new module for automatic calibration of hydrological parameters is now available. The module uses the SCE-UA algorithm, specially developed for this objective. Update your RS MINERVE software to try this new feature.

New indicators for model performance assessment

New indicators have been added in the Comparator object for assessing the performance of the hydrological simulations compared with observations.
Seven indicators are now available: Nash, Nash-ln, Pearson correlation coefficient, Bias Score, Relative Root Mean Square Error, Relative Volume Bias and Normalized Peak Error.
You can find more information about these indicators in the Technical Manual.

New User’s Manual with examples

Examples of application in new version of the RS MINERVE – User’s Manual  on the page Documentation.

New version of the User’s Manual includes some examples of application for new users. Enjoy practizing with them!