The UAE Cloud Seeding Program: A Statistical and Physical Evaluation

• Published in: Atmosphere Vol. 12; no. 8; p. 1013
• Main Authors: Al Hosari, Taha, Al Mandous, Abdulla, Wehbe, Youssef, Shalaby, Abdeltawab, Al Shamsi, Noor, Al Naqbi, Hajer, Al Yazeedi, Omar, Al Mazroui, Alya, Farrah, Sufian
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 01.08.2021
• Subjects: Aerosols; Aircraft; Annual rainfall; Archives & records; Arid regions; Arid zones; Atmospheric precipitations; change point; Climate change; Cloud seeding; Clouds; Decision making; Evaluation; Experiments; Gauges; Meteorological radar; Meteorology; Physical analysis; Precipitation; Radar; Rain; rain gauge; Rain gauges; Rainfall; Rainfall trends; Records; regression; Regression analysis; Statistical analysis; Statistics; Storms; Time series; Weather radar; United States > US; Israel
• ISSN: 2073-4433; 2073-4433
• DOI: 10.3390/atmos12081013

Operational cloud seeding programs have been increasingly deployed in several countries to augment natural rainfall amounts, particularly over water-scarce and arid regions. However, evaluating operational programs by quantifying seeding impacts remains a challenging task subject to complex uncertainties. In this study, we investigate seeding impacts using both long-term rain gauge records and event-based weather radar retrievals within the framework of the United Arab Emirates (UAE) National Center of Meteorology’s operational cloud seeding program. First, seasonal rain gauge records are inter-compared between unseeded (1981–2002) and seeded (2003–2019) periods, after which a posteriori target/control regression is developed to decouple natural and seeded rainfall time series. Next, trend analyses and change point detection are carried out over the July-October seeding periods using the modified Mann-Kendall (mMK) test and the Cumulative Sum (CUSUM) method, respectively. Results indicate an average increase of 23% in annual surface rainfall over the seeded target area, along with statistically significant change points detected during 2011 with decreasing/increasing rainfall trends for pre-/post-change point periods, respectively. Alternatively, rain gauge records over the control (non-seeded) area show non-significant change points. In line with the gauge-based statistical findings, a physical analysis using an archive of seeded (65) and unseeded (87) storms shows enhancements in radar-based storm properties within 15–25 min of seeding. The largest increases are recorded in storm volume (159%), area cover (72%), and lifetime (65%). The work provides new insights for assessing long-term seeding impacts and has significant implications for policy- and decision-making related to cloud seeding research and operational programs in arid regions.