Projection of precipitation and temperature in major cities of Pakistan using multi-model ensembles

Abstract

This study projects future variations in monthly precipitation and average temperature in major cities of Pakistan. It constructs Multi-Model Ensembles (MMEs) by selecting 16 General Circulation Models (GCMs) from the Coupled Model Intercomparison Project Phase 6 (CMIP6). In this process, it utilizes an Artificial Neural Network-based statistical downscaling approach. The performance of these models was assessed using five statistical metrics: Correlation Coefficient, Nash–Sutcliffe Efficiency, Root Mean Squared Error, Kling–Gupta Efficiency, and the Modified Index of Agreement. The results show that MMEs outperform individual GCMs in simulating historical temperature and precipitation trends across the cities. Projections for 2024–2100, based on four Shared Socioeconomic Pathways (SSP1–2.6, SSP2–4.5, SSP3–7.0, and SSP5–8.5), reveal a decline in annual precipitation by 39.22 %, 48.79 %, 36.27 %, and 38.08 %, respectively. In terms of temperature, maximum temperature is projected to rise by 5.95 % (+1.85 °C), 12.79 % (+3.97 °C), 9.86 % (+3.06 °C), and 16.22 % (+5.04 °C), while minimum temperature is projected to decrease by 4.25 % (−0.76 °C) and 0.74 % (−0.13 °C) under SSP1–2.6 and SSP2–4.5, respectively. However, under SSP3–7.0 and SSP5–8.5, the results show that minimum temperature is expected to increase by 0.20 % (+0.04 °C) and 7.26 % (+1.30 °C), respectively. The greatest potential for precipitation decline is seen in Islamabad, Multan, and Sialkot. At the same time, higher increases in maximum temperature are expected in high-altitude cities like Quetta and Peshawar compared to low-altitude areas. This study provides essential insights to help urban policymakers and stakeholders develop targeted policies for addressing the impacts of climate change in cities.