Thesis subject

MSc thesis topic: Identifying flash droughts with multimodal satellite remote sensing

Flash droughts are impactful climatological phenomena that are becoming increasingly recognised and researched. The characteristics of flash droughts, including sudden onset, rapid evolution, and severe impacts on agriculture and society, make their identification of considerable importance.

Due to their rapid development there is less early warning for impact preparation, whereas the common short duration means they are often not registered in the available drought monitoring products, such as the “Brazilian Drought Monitor” that reports on a monthly basis.

Over the years, a number of different remote sensing approaches have been developed and successfully applied for the monitoring of conventional droughts. However, the temporal resolution at which these are applied restricts identification of flash droughts. This project will explore whether sub-monthly time series analysis of metrics extracted from satellite optical and thermal imagery can enable detection of flash droughts in a semi-arid climate. The focus area will be Ceará state in north-eastern Brazil where flash droughts occurring during the crop growing season (referred to locally as “veranicos”) lead to significant losses in rainfed crop yields.

Relevance to research/projects

This project is aligned with the “Diagnosing drought for dealing with drought in 3D” project led by the Water Resources Management (WRM) group. As such, the supervision will be shared between the GRS and the WRM.

Objectives

  • Review of literature on remote sensing detection of drought onset
  • Time series analysis of metrics extracted from thermal and optical imagery
  • Comparison against ground rain gauge records

Literature

Requirements

  • Experience in using satellite remote sensing data
  • Good knowledge in scripting is a preference (e.g. Google Earth Engine, R, Python)

Theme(s): Sensing & measuring; Integrated Land Monitoring