|Dec 3, 2012|
|8 years 1 month ||Adding a check for nodata in the asserts on density
|Dec 2, 2012|
|8 years 1 month ||Sorry for a nasty typo!
|8 years 1 month ||While hunting for an alleged precipitation rate error (??, see
comment to issue 284), the non-updated;-) asserts for density
revealed a minor error regarding element removal: elements with
(almost) no more mass (M<Constants::eps2) could be joined to…
|8 years 1 month ||Update to discard MS_TOTALMASS from output for SN_GUI. Relocate HS1
input data accordingly to open position (50)
|Nov 30, 2012|
|8 years 1 month ||Tweaking the assert() in order to tolerate density=0 and temporary
slightly negative mass in sublimation (the element that will be
removed receives a mass that can be slightly negative because of
rounding errors, but will be deleted right away anywa…
|Nov 28, 2012|
|8 years 1 month ||Solving a bug when dealing with excess_water in the lowest element.
|8 years 1 month ||Now putting asserts on the element density (that must be positive)
|Nov 27, 2012|
|8 years 1 month ||Constification and asserts() put on elements mass. Every time the
mass of an element is changed, an assert checks that the mass
remains positive. The macro NDEBUG is already defined in cmake, so
when compiling in release, no extra cost is incured.
|Nov 26, 2012|
|8 years 1 month ||Some code cleanup (putting code in a function) and fixed a potential
problem when initializing the Sun object (when the position was only
given by the sno file, the object was not properly initialized).
|8 years 1 month ||The level of warnings when compiling in debug mode is now much
higher. Therefore the ProcessDat structure had to be adjusted as
well as solver.h. The external dependencies are now included as
system headers in order to avoid producing warnings of th…
|Nov 5, 2012|
|8 years 2 months ||Changing key R_IN_N into HARDNESS_IN_NEWTON
|Nov 4, 2012|
|8 years 2 months ||Introduce a config key pair "R_IN_N"/"OUTPUT"
that allows the user chosing to output hand hardness R either in
newtons or as index (steps).
By default R_IN_N is set to true. The io.ini of snowpack-opera sets
this to false.
|Nov 2, 2012|
|8 years 2 months ||One more compilation round with the DEEP_WARNINGS and then with
Weffc++. This lead to initializations in the initialization lists
and a copy constructor for SnowStation (the compiler was begging for
|Nov 1, 2012|
|8 years 2 months ||A few typos, partly to comply with doxygen ;-)
|8 years 2 months ||The stability parameters in classes SnowStation, Stability, and
ProcessDat(Ind) needed more thorough checks to avoid trying to
insert big numbers in SDB. By the same token the catching up of
faulty data records into SDB has been improved
|Oct 17, 2012|
|8 years 3 months ||Some constification in compTemperatureProfile and added a new field
to CurrentMeteo (dir_h, the direct radiation on the horizontal, to
be used in a later commit).
|8 years 3 months ||After the new energy balance solver, a convergence problem for the
temperature equation was encountered. This is solved by increasing
the allowed maximum number of iterations. Maybe the situation should
get a closer look, because something else may b
|Oct 6, 2012|
|8 years 3 months ||In version 57 Fierz made a tremendous error at attempting to use
(date1 - date2).getJulianDate() in Hazard.cc to compute hns
values. This was not noticed until Ulrich Niederer tried to compute
threefold sums of hns3 values!!!
Now it is back to orde
|Oct 4, 2012|
|8 years 3 months ||A follow-up commit to remove an unused variable.
|8 years 3 months ||This commit is an important change to SNOWPACK to strongly improve
energy balance. In melt conditions, the temperature of the surface
doesn't change. This is something which is not described by the
temperature equation. Therefore, in melting conditio