The Nature of the Beast

For the last two weeks I've been comparing my data, to that of my predecessor's (both lab and simulated) without too much success.
If I do a bit of a rewind here, my Phd revolves mainly of modelling pollutant travel through urban drainage volumes. Think Manholes; Tanks Zzzzzzz.
I've got data from a scale model, a CFD model and some fancy PIV data (cameras that work out magnitude and direction of particles within flow!) for verification against what is meant to be my new and improved model to be used in more realistic (i.e. changing) conditions.
The fit was appalling and today I found out why. I traced it all back to particles going missing. Numerically vanishing. At first it seemed completely random but after a LOT of coding and analysing of data. Think 1e+9 lines of data per particle for 80,000 particles. I found out where they were coming from and where they were vanishing.
I've been convinced that I'm a dim wit and got something massively wrong in the setup of all my models.
I've been chasing a numerical glitch in a commercial code for at least the last month. The thing is; during the past month I've cursed my work, my supervisor, my project and a lot of other things yet today I have an almighy grin on my face as I've triumph'd, this is the nature of the beast. Problems challenge you to devlop mentally, examine things you might otherwise ignore and learn new skills to bust the problem wide open.

BOOM