A framework for climate change assessment in Mediterranean data-sparse watersheds using remote sensing and ARIMA modeling

• Published in: Theoretical and applied climatology Vol. 143; no. 1-2; pp. 639 - 658
• Main Authors: Al Sayah, Mario J., Abdallah, Chadi, Khouri, Michel, Nedjai, Rachid, Darwich, Talal
• Format: Journal Article
• Language: English
• Published: Vienna Springer Vienna 2021; Springer; Springer Nature B.V
• Subjects: Adaptation; Analysis; Aquatic Pollution; Aridity; Atmospheric models; Atmospheric Protection/Air Quality Control/Air Pollution; Atmospheric Sciences; Autoregressive models; Climate change; Climate change mitigation; Climate models; Climate science; Climate trends; Climatic analysis; Climatic data; Climatic evolution; Climatology; Confidence intervals; Drought; Droughts; Earth and Environmental Science; Earth Sciences; Environmental assessment; Global temperature changes; Indicators; Landsat; Landsat satellites; Mediterranean climate; Mitigation; Original Paper; Precipitation; Precipitation variability; Remote sensing; Satellite imagery; Statistical analysis; Trend analysis; Waste Water Technology; Water Management; Water Pollution Control; Watersheds; Weather forecasting; Lebanon
• ISSN: 0177-798X; 1434-4483
• DOI: 10.1007/s00704-020-03442-7

This study aims to propose a framework for assessing climate change in Mediterranean data-sparse contexts. For that purpose, the 309-km 2 Lebanese Nahr Ibrahim watershed, extending over 3% of Lebanon’s surface, was chosen as a representative of the targeted settings. Generally, holistic climate change assessments encompass both climate trend analysis and future forecasting. According to the World Meteorological Organization, a continuous, homogenous, and uninterrupted climatic record for at least 30 years is needed to fulfill these tasks. Often, some Mediterranean watersheds lack such data and are hence characterized by climatic data scarcity. Such is the case of Lebanon where 30 years of wars have considerably disrupted the country’s climatic record. In an effort to overcome this state of data scarcity, remote sensing–derived drought indicators were used to determine the climate’s evolution during the last 28 years. For that purpose, several remote sensing indices were extracted from LANDSAT imageries for the period 1990–2018 at a 3-year interval, and were coupled to meteorological indicators. Forecasting was then performed using autoregressive integrated moving average (ARIMA) models. Meteorological indices showed increased variability of precipitations and aridity periods, while remote sensing indicators collectively revealed slight shifts towards increasing droughts. Projections using ARIMA models forecasted increases of 0.9 °C, 0.7 °C, and 0.8 °C for average, maximal, minimal temperatures, and an average 6 mm decrease of precipitations at the 95% confidence level for the year 2030. The presented approach can serve as a tool for proactive climate change mitigation or adaptation plans.