In preparing the protocols for the upcoming AstroPlant implementation in Gent, defining how the sensors generate data and measure the environment, how the students are supposed to work with the kit, we had a few questions and decided to ask Christophe Lasseur, the director of the MELiSSA consortium, who recently received a honorary doctorate for scientific merit from the University of Antwerp’s Faculty of Science. We summarise these questions and answers below.
Photo credit: ESA Images
With this data, what will be needed, for example, for each of the data points, what should be the frequency and accuracy?
Temperature of the air: if automated, hourly is enough. If manual: twice a day
Humidity: same as temperature
CO2 level: several times per hour. Every 5 min is good
Light: once daily (not when light start but when they have been on for an hour at least) and each time the light is changed through a dimmer (if there is a dimmer). No need of more, especially if the light intensity and spectrum are evaluated beforehand.
Temperature of the water: once a day like EC
EC of the water: once a day is fine
pH of the water (manually): once a day
Images of the plant (camera): in visible light, one during day time and one during night time (whenever possible); in IR light, hourly during day time, once during night time
What are the specific scientific and practical goals behind the plant characterisation process? Is it optimising nutrient uptake? Experiments with lighting? Temperature? What are you trying to optimise for?
The idea is really to collect data for modelling. So plant size, leaves size, leave organisation, leaves surface, water balance, will be important. We are trying to get data for modelling of the morphological characteristics of the plants, we are not trying to optimise any of the environmental parameters.
Which plants do we start with?
We propose to start with lettuce (kropsla Austina or Grand Rapid) and spinach rather than basilicum.
Do we want the citizen scientists to manually collect data about leaf size, weight (after growth), stem diameter and length, number of branches, root network, data about the fruits (if applicable)?
Manual data collection is necessary as a way to engage the user and cross-check the data, which can be acquired from the sensors measurements. Particularly the data acquired by imagery for which accuracy (I certainly mean good calibration here) is needed. The issue we see with the manual data collection are :
1- metric reference,
2- exact definition of the object to be measured (which dimensions, from where to where…),
3- exact definition of how to measure.
We may need both but manual may induce plant pathogens too. TBD.