Wednesday, 6 February 2013

Getting into Carbon Science

Obtaining estimates of Carbon emissions through the method of inverse modelling was not something I had in mind for myself five years ago. I got into the CSIR with a background in mathematical statistics and ecology. I had just finished a three year lecturing position, which I did while completing my Masters in stats, at a nearby university (attempting to teach statistics to first year biology students – fun), and couldn’t wait to start my career in pure research. Best of all, the group I was joining was a systems ecology group, and so I was going to get to do work in both of my areas of interest.

I started off with processing and modelling data from our Skukuza eddy-covariance flux tower. The data collected here provides information on carbon dioxide and water fluxes at a landscape level. This helps us to better understand processes like primary productivity and ecosystem respiration, as well as evapotranspiration. At the same time meteorological variables like air temperature, relative humidity, wind speed and direction, and net radiation are measured, as well as soil variables, like moisture and temperature. This helps us to understand what the drivers are, and also to predict what will happen with ecosystem productivity and respiration under different climate change scenarios.

Skukuza Flux Tower – Kruger National Park
See for more information on eddy covariance methodology 
I was minding my own business one day in the office, doing fun and exciting things in R, when my boss called me up and asked me if I was interested in instrumentation and getting things to work. Looking back now, this was a turning point in my life. If I knew then what I know now, I’m not so sure I would have answered yes. He handed me an old broken PP Systems carbon dioxide gas analyser, specifically for measuring soil fluxes, which looked to me like it had been built back in the day of the first black and white TV, and basically told me to get it to work. At university, when you’re studying statistics or ecology, nowhere are you taught any electronics courses. That’s the sad reality of it. So I was deeply overwhelmed, but fortunately got given the name of a very nice man who was at the manufacturing unit, and was a whizz with a soldering iron, and he helped me to figure out what the problem was. Really – he figured out what the problem was and I watched. That was my first experience of getting down and dirty with measuring carbon dioxide. The next thing I found myself attending a summer school for flux measurements held at The University of Colorado Mountain Research Station. I should have known it wasn’t just so that I could have fun in Colorado.

It wasn’t too long before I was asked to write a program for a datalogger to collect information from a soil moisture probe. This basically entails having to understand how the instrument works, and what instruction to give it to work, and how to interpret the output of the instrument, which is usually voltage, and turn it into a measurement. I’m always amazed every time I wire an instrument up to a datalogger, and it starts giving out actual measurements. It leaves one with a real sensation of satisfaction when the data all looks good. On the other hand, when you wire up an instrument exactly the way the instructions tell you to (or you think you have), and it gives out garbage, that’s an entirely different sensation. Frustration, anger, resentment, disappointment, despair, hopelessness …

This may look like a mess of wires to you, but each of those has its place and needs to correspond exactly to the program written to collected the data
 Before I knew it, I was writing a program to get an entire flux tower and all its meteorological instruments to operate. And not long after that, I was finding myself in the field, attending to the every whim and desire of the instruments. We have two flux towers in the Kruger National Park. When things are working, life is great and the world is a happy place. But sometimes you can arrive on site, and be inundated by every manner of problem that can befall an instrument. Everything from elephants pushing a tree over onto the solar panel, ants squatting in the instrumentation box and causing all sorts of shorts, or in fact gecko’s taking up house – or scratch that – an entire ecosystem existing in the instrumentation box. You’ll also be surprised how much damage the African sun can do to instrument cables. Many of the instruments that we use are shipped in from America or Europe, and having to deal with the kind of UV we have down south here doesn’t always come into the equation. Don’t factor out the wind and rain either. And then there’s human error, which is the kind which causes me the most frustration and sleepless nights. I’m particularly good at making those, and I’m always gobsmacked by how spectacularly wrong things can go when just a tiny little detail is over looked.

Repositioning of the foliage by the local elephant hooligans
 Let me tell you … working with instrumentation is not for the faint of heart.

No comments:

Post a Comment