Statistical analysis of stream discharge in response to climate change for Urumqi River catchment, Tianshan Mountains, central Asia

Abstract

Understanding the impact of climate change on water resources is essential to sustainable development. Urumqi River, a medium-size headwater catchment in the Tianshan mountains, was chosen to evaluate the impact of climate change on its discharges, where impacts of human activities on water are negligible and the influence of glacier melt is minor as far as the mountainous outlet is concerned. Analysis of the time series of temperature, precipitation and river discharge of the hydrological station at the mountainous outlet from 1959 to 2006 was carried out using different statistical methods, including linear regression, Mann–Kendall test and wavelet analysis. Although both temperature and precipitation show a significant upward trend with a gradient of 0.02 °C/y and 2.08 mm/y, respectively, there is no significant rising trend in the stream discharge. The reasons are attributed to the hysteresis and buffering effects of groundwater in conveying the change from precipitation to stream discharge. Common short periods of less than 8 years exist in the three time series, which is a common trend in NW China. Precipitation has stronger influence on stream discharge than air temperature throughout the 48 years of instrumental records in the Urumqi River catchment.

Introduction

Climate warming during the last century has been recognized as a fact supported by both theoretical analysis and instrumental data (IPCC, 2007; Gregory et al., 2012). Changes in air temperature and precipitation have significant impacts on hydrological processes and water resources. Previous studies have mostly focused on the impact of climate change on river discharge to which glacier-melt water contributes a substantial part (Lafreniere and Sharp, 2003; Labat et al., 2004; Li et al., 2010) while few have done for non-glacier dominated catchments. Furthermore, human activities exacerbate the complexity of stream behavior through altering the natural recharge, flow and discharge. It is difficult to differentiate the individual impact of climate change from that of human factors on water resources (Jones, 2011; Tao et al., 2011). In order to avoid such ambiguity, an alternative method is to study river catchments in regions without human disturbance so as to focus on the single impact of climate change on water resources (Kong and Pang, 2011). Headwaters in alpine regions are one of the most appropriate research subjects for studies in this regard. This is because (1) alpine headwaters are almost free from the effects caused by human activities and (2) they are constituted of glacier-melt water, which is quite sensitive to temperature change. A wide variety of recent studies addressed this issue and examined the significance of climate change on water resources (Chen et al., 2006, 2008; Immerzeel et al., 2010; Li et al., 2011, 2010; Xu et al., 2011a, Xu et al., 2011b).

Urumqi River, located in an alpine region in Northwest China, is a good example of this kind. Previous studies found that at Glacier No.1, which is the source of Urumqi River, discharge increased greatly due to the rise of temperature and precipitation from 1959 to 2006 (Li et al., 2007, 2010). Discharge of Urumqi River at Yingxiongqiao hydro-meteorological station, which is its mountainous outlet, is mainly composed of groundwater, with less than 9% glacier-melt water (Kong and Pang, 2012). Thus, the change of mountainous discharge and its response to climate change must be different from that at the source (Glacier No.1). The mechanism of Urumqi River's response to climate change could be comparable to other rivers that are not significantly affected by glacier-melt.

This paper examines the trend of change of climate and discharge of Urumqi River catchment and the interrelationship between them. Several statistical methods including linear regression, Mann–Kendall test, and wavelet analysis are used to delineate the changes, aiming at: 1) to detect any climate change effects in the time series of temperature, precipitation and discharge; 2) to identify possible co-variability of the three parameters; and 3) to assess impact of climate change on stream discharges at different time scales.