Aggregated drought characteristics using particle swarm optimisation incorporating duration, severity, reliability, resilience, and vulnerability characteristics

ABSTRACT A deeper understanding of drought risks is presented for the basin of Lake Urmia, where the Lake disappeared catastrophically during the years 1995–2023. Received wisdom attributes the catastrophe to droughts and/or climate change, but the authors identify more than 40 dams in this small basin as the singular most likely cause, which has sprung up since 1995. The article adds further evidence by studying droughts through a modelling strategy comprising the following: (i) use the standardised precipitation index (SPI) to map SPI at 1, 3, 6, and 9 months timescales with 21 years of recorded data; (ii) derive duration and severity (DS) values at 55 observation stations; (iii) derive reliability, resilience, and vulnerability (RRV) indicators using DS; (iv) aggregate the subsequent five maps (DS + RRV) for the four timescales using a particle swarm optimisation algorithm to compact their inherent information. The aggregated results show: (i) 1-month timescale: wet and dry zones are demarked by an axis running along northwest-southeast of the basin; (ii) 3-month timescale: the aforementioned behaviour prevails, but the axis runs along east-west; (iii) higher timescales: the wet and dry zones flip. The overall results show that droughts are natural features of the basin but not catastrophic.