PhD: Using data assimilation and weather forecasts in irrigation scheduling

 (via EURAXESS)
KU Leuven
Leuven, Belgium
Position Type: 
Scholarship
Organization Type: 
University/Academia/Research/Think tank
Experience Level: 
Not Specified

EXPIRED

Please note: this job post has expired! To the best of our knowledge, this job is no longer available and this page remains here for archival purposes only.

With the development of cheap and autonomous soil sensors that transmit information about the soil directly and continuously to data servers, farmers may have now access to unprecedented real-time information about the state of their fields. An important challenge is to use this information effectively and improve the revenues while reducing the production costs and pressures on the environment. In irrigated agriculture, this challenge can be translated to a more effective water use to increase and improve yields and reduce water losses. A more effective water use comes down to applying the right amount of water at the right times. 

In Flanders, the Bodemkundige Dienst van België (BDB) already offers irrigation scheduling for farmers as a paid service.  The advice is based on simulations with a soil water balance model fed with weather data in real time, and the model is calibrated with monthly soil moisture data from (manual) soil sampling. This could be improved by also using weather forecasts and by using inexpensive soil moisture sensors that send their data via Internet-of-Things (IoT) technology to a server that runs model-based irrigation advisory system.

But, both weather forecasts and the properties of the soil and crop as used in the model are uncertain. These soil and crop properties are estimated from analysing previous information about how the soil moisture changed over time. This analysis and estimation of the soil and crop properties or model parameters corresponds with ‘training’ the model so that it can reproduce the observed soil water content dynamics (sensor data). Data assimilation is a technique that can be used to train a model. Since it also acknowledges uncertainties in the information that is used (the soil moisture and crop observations and the uncertainties in the weather forecasts), the uncertainty in the forecasts of the soil moisture, lost water from the root zone and crop water water stress can also be evaluated. The technique is promising for improving irrigation scheduling, but to the best of our knowledge not yet used in irrigation advisory systems.

The PhD research will contribute to a VLAIO-funded 4-year research project on (supplementary) drip irrigation for vegetable crops in Flanders. The objective of the project is to develop an irrigation scheduling tool for vegetable production in Flanders, and to tackle a few technical constraints for drip irrigation in Flanders.