Developing objective dry spell and drought triggers for drought monitoring in the Niger Basin of West Africa

To effectively manage drought in the Niger River Basin known for its high interannual rainfall variability and continuous adverse drought impacts and challenges, decision-makers need to distinguish dry spells from drought. Operational drought triggers also needed to activate water resource management measures and determine the level of intervention, which are presently lacking. With 36 years of precipitation reanalysis records from 60 locations, the study attempts to establish index- and percentiles-based thresholds for defining dry spell and drought through baseline assessment analysis of the historic 1980s regional drought-induced famine, using percentile rank approach. Results indicate 1980s drought-induced famine occurred within thresholds of 20th, 10th, 5th, and 2nd percentiles, while the drought precursors, the dry spells occur within 35th, 20th, 10th, and 5th percentiles. The corresponding objective index thresholds based on SPI-2 month ranges from − 0.22 to − 0.45, − 0.45 to − 0.93, − 0.93 to − 1.20, − 1.66 to − 1.83 defining dry spell, abnormal dry spell, critical/extreme dry spell, and drought conditions; and SPI-6 month thresholds of range − 0.38 to − 1.07, − 0.59 to − 1.58, − 0.64 to − 1.79 and − 0.67 to − 2.21 defining drought of moderate, severe, extreme, and exceptional intensities respectively depending on location. The thresholds vary from month to month, reflecting seasonality, and detect drought onset much earlier because of its lower values relative to the higher, subjective, and arbitrary fixed SPI thresholds of − 1.0 to − 1.49 for defining moderate drought conditions that could be misleading. The threshold validation results show a success rate of 50 to 70 percent. Results operationally useful for early drought detection.