Status and Prospects of Drought Forecasting
Professor Amir AghaKouchak, University of California, Irvine, USA
Improving water supply management and decision making in water stressed regions requires reliable seasonal drought prediction, which remains a grand challenge. Over the past decades, a wide range of models, including dynamical, statistical and hybrid dynamical-statistical, have been developed for seasonal precipitation forecasting. However, each type has its own advantages and disadvantages. This presentation offers a review of progress in different types of forecasting methods including numerical/dynamical (eg, North American Multi-Model Ensemble), statistical models (eg, analog-based, ensemble streamflow prediction, artificial intelligence), and hybrid approaches that combine the two. This talk will review different hybrid frameworks such as hierarchical and parallel model concepts. The presentation ends with a discussion on current gaps and opportunities including satellite observations and artificial intelligence techniques.
Recent advances in drought monitoring using satellites and models
Professor Justin Sheffield, University of Southampton, UK
UK droughts past, present and future: quantifying hydroclimatic variability to inform drought management
Associate Professor Jamie Hannaford, UK Centre for Ecology & Hydrology and Maynooth University, UK and Ireland
Water resource managers need to understand hydroclimatic variability to manage droughts. For long-term planning it is essential to understand drought risk, based on past recorded variability and what is likely to happen in future. For managing drought events in the here and now, it is necessary to understand current water availability status, and what is likely to happen over the coming weeks and months. While the UK has a long history of water resources management and well-evolved frameworks for drought planning, at the same time there are inevitable limitations: our understanding of the past is constrained by short observational records, while the utility of future forecasts and projections (whether seasonal or multi-decadal) is inevitably constrained by the uncertainties involved in climate and hydrological modelling. This presentation focuses on recent advances in our understanding of drought variability, largely drawn from the UKRI Drought and Water Scarcity Programme (2014 – 2020). This will include efforts to extend our understanding of past drought variability through hydroclimatic reconstructions, initiatives to provide national-scale, spatially coherent hydrological projections for the 21st century, and the latest advances in status monitoring and seasonal forecasting. Several key themes run through all this work: notably the importance, and challenges, of drought definition and quantification; and the importance of co-development with diverse end-users to ensure the applicability of hydroclimate services in drought management on-the-ground.
Benchmark worst droughts in India (1901-2020)
Dr Vimal Mishra, Indian Institute of Technology (IIT) Gandhinagar, India
During the summer monsoon (June-September) season, drought poses challenges for agricultural activities and water availability in India. We develop a framework considering the timing, areal coverage, and severity of droughts that can be used for the assessment as the monsoon season progresses. We estimate the benchmark worst droughts within the monsoon season (June, July, August, and September) using the long-term (1901-2020) gridded rainfall. The benchmark worst droughts were identified considering the extent and severity of drought using the Drought Severity Coverage Index (DSCI). The worst meteorological droughts in June, July, August, and September occurred in 1923, 2002, 1937, and 1907 with a return period of 68, 200, 147, 188 years, respectively. The worst drought in the entire summer monsoon season occurred in 1918, which had a return period of 238 years. The benchmark droughts during June 1923, July 2002, and monsoon 1918 were associated with the warm SST over the equatorial Pacific Ocean. The other two droughts (1937 and 1907) were linked with the off-equatorial warming over the Indo-Pacific region. The estimated DSCI for a 2-500 return period can be used for the drought assessment during the monsoon season in India.