Weather observations are typically collected by professional devices siting in open spaces. These weather stations provide very precise measurements, but they also require a regular maintenance and calibration of the sensors, which implies a high cost for national weather services. In general terms, one could say that we are good with the current weather forecast. In the end, the combination of in-situ measurements with numerical weather models and remote sensing data, provides a decent weather forecast for several days in advance.
Well, another PhD journey comes to an end. These have been fruitful and intense years of research and fast-paced learning. The PhD research is one of the few occasions in one’s professional career in which you have the opportunity of giving full dedication to study a topic in depth for several years. The result of this investigation are four papers that, hopefully, will help and inspire new researchers to apply machine learning to model elusive and fine-grained phenomena, and to be enthusiastic and open-minded about using citizen science data to animate new research lines.
The fourth paper is currently under review for PLOS ONE, but you can read the pre-print in bioRxiv. In this work we basically take all the knowledge gained and the building blocks developed in the three previous publications to create a tick bite risk model. Let’s recap what we know so far:
Risk ®: R can be estimated from the volunteered tick bites, which are a combination of H x E R seems to be more influenced by human factors than seasonal accumulations of weather variables Hazard (H): H can be estimated from the tick activity counts collected by volunteers in forested locations H seems to be driven by atmospheric water levels, rather than temperature Exposure (E): E levels are similar in unattractive suburban forest patches E levels are maximum along the edges of attractive forests and big natural areas In the third publication we show how obtain a static map of human exposure as a combination of the other two components.
The third paper associated with this research was published in Scientific Reports in 2018. Let’s first recap the two previous publications. In the first one we learnt that the volunteered tick bites collection represents the risk ® of getting a tick bite. In the second paper, I explained how to obtain daily tick activity, which is a proxy of the tick hazard (H) of a location. However, there is a third component in the equation: human exposure.
The second paper of my research was published in 2017 in the International Journal of Health Geographics. In this article we explain what is our approach to model tick activity, a proxy for tick hazard (H). Tick activity is a sneaky phenomenon to model, because it is the product of several simultaneous natural phenomena such as weather, vegetation changes, or wildlife dynamics (not available in this research). These factors are intertwined with each other and they determine tick survival, by creating (un)favorable conditions for ticks to thrive/die throughout the year.
The first paper was published in 2016 in a special issue of Transactions in GIS on the Role of Volunteered Geographic Information in Advancing Science. A year before, a very enthusiastic and naïve self had received a collection of 35,000 volunteered tick bites provided by the RIVM. I had some experience working with volunteered data because in my previous research group, I worked in several projects using this type of citizen-contributed information.