When and how will the Millennium Silk Road witness 1.5 °C and 2 °C warmer worlds?

Abstract Western China and central Asia are positioned centrally along the Millennium Silk Road, which is regarded as a core region bridging the East and the West. Understanding the potential changes in climate over this core region is important to the successful implementation of the so-called ‘Belt and Road Initiative’ (a $1 trillion regional investment in infrastructure). In this study, both mean and extreme climate changes are projected using the ensemble mean of CMIP5 models. The results show a warming of ~1.5, 2.9, 3.6, and 6.0 °C under RCP2.6, 4.5, 6.0, and 8.5, respectively, by the end of the twenty-first century, with respect to the 1986–2005 baseline period. Meanwhile, the annual mean precipitation amount increases consistently across all RCPs, with an increase by ~14% with respect to 1986–2005 under RCP8.5. The warming over the Millennium Silk Road region reaches 1.5 °C before 2020 under all the emission scenarios. The 2020s (2030s) see a 2 °C warming under the RCP8.5 (RCP4.5) scenario. Global warming that is 0.5 °C lower (i.e. a warming of 1.5 °C) could result in the avoidance of otherwise significant impacts in the Silk Road core region—specifically, a further warming of 0.73 °C (with an interquartile range of 0.49%–0.94 °C) and an increase in the number of extreme heat days by 4.2, at a cost of a reduced increase of 2.72% (0.47%–3.82%) in annual precipitation. The change in consecutive dry days is region-dependent.


Introduction
The Silk Road is a historically important international trade route between China and the Mediterranean. As a network of trade routes, formally established during the Han Dynasty of China around 130 BCE (before common era) to open trade with the west, the Millennium Silk Road has served as a thoroughfare from east to west and linked the regions of the ancient world in commerce. Because China's silk comprised a large proportion of the trade along this ancient road, in 1877 it was named the 'Silk Road' global mean during recent decades has been found for Northwest China (Hu et al. 2014;Li, Chen, and Shi 2012). Meanwhile, associated with the increased specific humidity and southward displacement of the Asian subtropical westerly jet, a significant wetting trend has been seen in this region (Peng and Zhou 2017). Many factors have been suggested to affect the climate of central Asia, including the interannual variability of the Asian westerly jet , El Niño-Southern Oscillation (Mariotti 2007), the Indian summer monsoon , and circumglobal teleconnection (Ding and Wang 2005;Huang et al. 2015). These studies generally focused on interpreting observed climate changes, whereas less effort has been devoted to climate change projection Wang et al. 2017), as compared to the many publications on the projection of East and South Asian climate change (Chen, Xu, and Yao 2015;Chen and Zhou 2015;Chen and Zhou 2017;Guo et al. 2016Guo et al. , 2017He and Zhou 2015;Kitoh et al. 2013;Li, Zou, and Zhou 2017;Peng et al. 2016;Zhou 2015, 2016).
The parties of the United Nations Framework Convention on Climate Change (UNFCCC) signed the Paris Agreement in December 2015. The Paris Agreement stated a long-term goal for climate protection as 'holding the increase in the global average temperature to well below 2.0 °C above pre-industrial levels and pursuing efforts to limit the temperature increase to 1.5 °C above pre-industrial levels, recognizing that this would significantly reduce the risks and impacts of climate change. ' (UNFCCC 2015). As a result of the Paris Agreement, there is considerable interest in the scientific community to provide input on the projected climate impacts under a climate warming scenario that stabilizes at 1.5 °C, in terms of water availability, temperature and precipitation extremes, as well as ecological and agricultural impacts (Hulme 2016;King, Karoly, and Henley 2017;Li et al. 2018;Mitchell et al. 2016;Schleussner et al. 2016aSchleussner et al. , 2016b. For China specifically, extreme precipitation days (annual number of days with precipitation greater than 50 mm) and total precipitation during days when daily precipitation exceeds the 99th percentile, are projected to increase by 25.81% and 69.14%, respectively, with respect to the 1971-2000 base period, as determined by model ensemble data from phase 5 of the Coupled Model Intercomparison Project (CMIP5; Guo et al. 2016). The Intergovernmental Panel on Climate Change (IPCC) has accepted an invitation from the UNFCCC to provide a special report on the impacts of global warming of 1.5 °C above pre-industrial levels. However, information on climate change over the regions of the Millennium Silk Road in a 1.5 and 2 °C global warming world remains quite limited. Accordingly, this study aims to answer the following questions: (1) How will temperature and precipitation change over the SRC region in the coming century under by the German geographer and traveler Ferdinand von Richthofen. Western China and central Asia are positioned centrally along the Silk Road (Figure 1), and together are regarded as the Millennium Silk Road core region (hereinafter, the SRC region), important in linking the Eurasian continent's east and west coasts. The SRC region, consisting of five provinces of Northwest China (Xinjiang, Qinghai, Ningxia, Shaanxi, and Gansu) and five central Asian countries (Kazakhstan, Uzbekistan, Kyrgyzstan, Turkmenistan, and Tajikistan) (Figure 1(a)), is dominated by arid and semi-arid climate (Figure 1(b)). The development of society and economy here relies heavily on water resources, and is one of the 'hot spots' in climate research identified as particularly sensitive to climate change. Reliable information on climate change in the SRC region is of crucial importance to the successful implementation of 'the Belt and Road Initiatives (The Silk Road Economic Belt and the 21st-century Maritime Silk Road is a development strategy proposed by Chinese Government that focuses on connectivity and cooperation between Eurasian countries.)' . However, to date, the availability of such information has been quite limited.
Recent decades have seen a wetting trend along with warming conditions in some regions of the Millennium Silk Road. Efforts have been devoted to understanding observed past changes. A warming trend larger than the different warming scenarios? (2) When will the Millennium Silk Road witness warming of 1.5 and 2 °C? (3) How will the mean state and extreme climate events change over the SRC region in a 1.5 and 2 °C global warming future?

Data and methods
The historical simulations and projections under different Representative Concentration Pathways (RCPs) from coupled climate models in CMIP5 (Taylor, Stouffer, and Meehl 2012) are analyzed. The models used in the construction of the analysis fields given in this study vary with the research focus, since we hope to use as many model results as possible to reduce the uncertainties. As summarized in Table   1, data from 24 models are used to calculate the regional average annual mean surface air temperature (SAT) and precipitation under all four RCP scenarios (RCP2.6, RCP4.5, RCP6.0, and RCP8.5), shown in Figure 2. The projected SAT under two RCPs (RCP4.5 and RCP8.5) is derived from 40 CMIP5 climate/earth system models (Table 1) and used to calculate the 1.5 and 2 °C threshold-crossing times shown in Figure 3. Taking 2 °C warming for example, this is defined as the SAT anomalies larger than 2 °C relative to the mean state during 1861-1890 in the historical simulation. To reduce the uncertainty in calculating the threshold-crossing time due to interannual variability, a nine-year running average is used to smooth the time series of SAT anomalies. Then, the 2 °C threshold-crossing time is defined as the Table 1. details of the cmip5 models. All the models are used to derive the 1.5 and 2 °c threshold-crossing time in Figure 4. the models in italic type provide projections under all four rcp scenarios, and are thus used in Figures 2 and 3. the models in bold type are used to derive the extreme indices in Figures 6 and 7, based on daily data availability.
note: a indicates that the full name of JAmstec, Aori, nies is "Atmosphere and ocean research institute (the University of tokyo), national institute for environmental studies, and Japan Agency for marine-earth science and technology". the 1.5 and 2 °C warming scenarios are aggregated from the nine-year windows centered on the years when 1.5 and 2 °C warming above pre-industrial levels (1861-90) is reached. The climate changes in a 1.5 or 2 °C warmer world are calculated for each model separately to derive multi-model ensembles.

Results
We begin the analysis from the projected changes of annual mean SAT and precipitation over the SRC region ( Figure  2). Analysis of the changes over the SRC region is carried out along with a comparison with the changes averaged over the global land area and entire globe (Figures 3 and  4). Warming over the SRC region under different emission scenarios generally follows that of the global mean, but the magnitude is larger (Figures 3 and 4; also see Figure  12.5 in IPCC AR5, albeit the numbers of models used are slightly different). By the end of the twenty-first century, the multi-model ensemble mean projects a warming of ~1.5, 2.9, 3.6, and 6.0 °C under RCP2.6, 4.5, 6.0, and 8.5, respectively, relative to the 1986-2005 baseline (Figure 2).
first year when the SAT anomalies reach 2 °C. The model outputs are remapped onto a 5° × 5° grid by bilinear interpolation for reliable large-scale information.
The data from the 24 models in bold type in Table 1 are used to derive the extreme indices presented later in the paper because daily data is only provided by these models. The standard definitions of the extreme indices are adapted from the Expert Team on Climate Change Detection and Indices (Zhang et al. 2011). Two indices are used in our analysis. The first is 'consecutive dry days' (CDD), which is defined as the largest number of consecutive days with daily precipitation less than 1 mm. The second is 'extreme heat event' , which is defined as the annual count of days when the daily maximum temperature is greater than 35 °C. The extreme indices are calculated on the native grids of the individual models and then regridded to a common resolution of 1° × 1° to derive multi-model ensembles.
In our analysis, to account for future deviations from the current climate, a common reference period of 1986-2005 is referred to as the baseline scenario. The results for  These warming magnitudes are around 0.2-0.8 °C greater than the global land average, and 0.5-2.0 °C greater than the global mean (Figures 3 and 4). The more severe the emissions scenario, the greater the difference in warming. Such an amplification of warming in dry regions is related to the positive land surface feedback, as reported in Huang et al. (2017).
The annual mean precipitation amount increases consistently across all RCPs, although the model spread is large (Figure 2(b)). The percentage increases are faster than the global average (Figures 3(b) and 4(b)). For example, the annual mean precipitation amount is projected to increase by ~14% with respect to 1986-2005 over the SRC region by the end of the twenty-first century, as compared to ~6.5% for the global land area and ~6.8% for the global mean, under RCP8.5 (compare Figure 2 But when will the Millennium Silk Road reach a 1.5 and 2 °C warming? The respective timings under different emission scenarios and the corresponding changes are shown in Figure 2. Also, more specifically, we show the spatial patterns of the 1.5 and 2 °C threshold-crossing time under the RCP4.5 and RCP8.5 scenarios over the Eurasian continent in Figure 5. For the multi-model ensemble mean, the 1.5 °C threshold-crossing time in the SRC region is projected to be reached before 2020, under all the emission scenarios. The 2 °C threshold is projected to be breached no later than 2030. Under RCP8.5, this could be as early as 2022. It is unsurprising to see an expected earlier threshold-crossing time under the higher emission scenarios. However, it should be noted that there is large spread among the  Crucially, although generally a warmer and wetter climate mean state is expected under the 1.5 and 2 °C global warming scenarios, the specific changes between a 1.5 and 2 °C warmer world are regionally dependent. For instance, a 1.5 °C warmer world induces a regional annual mean warming of 1.24 °C relative to the 1986-2005 base period over the SRC region, with an interquartile range of 0.92-1.63 °C. Compared with the 1.5 °C warmer world, however, the additional 0.5 °C under global warming of 2 °C leads to a further warming by 0.73 °C (interquartile range is 0.49-0.94 °C, similarly hereinafter) over the SRC region, which is much greater than the global mean level. The annual mean precipitation amount increases by 4.77% (2.31%-8.59%) under warming of 1.5 °C relative to the baseline, but the extra 0.5 °C under warming of 2 °C brings a further 2.72% (0.47%-3.82%) increase in precipitation over the SRC region. The spatial pattern features CMIP5 models in the projection of the 1.5 and 2 °C threshold-crossing times (not shown), which can be explained in terms of different climate sensitivities to anthropogenic forcing in different models .
The threshold-crossing time over the SRC region is earlier than that over the global land area or global mean (compare Figure 2(a) with Figures 3(a) and 4(a)). Under the RCP8.5 scenario, for example, a regional 1.5 °C (2 °C) warming over the SRC region is breached in 2010 (2022), whereas over the global land area and for the global mean the years are 2015 (2028) and 2025 (2039), respectively.

Summary and concluding remarks
Western China and central Asia are positioned centrally along the Silk Road, and are together referred to as the Millennium SRC region. This region encompasses arid and semi-arid climate and is thus sensitive to climate change. The Belt and Road Initiative is a sophisticated network of significant increases against model uncertainty between the two warming levels in the eastern part of the Silk Road ( Figure 6(f )).
Changes in climate extremes may have larger impacts than mean-state changes. But how will extreme climate events change over the regions of the Millennium Silk Road in a 1.5 and 2 °C global warming future? Based on Figure 7, we examine the multi-model ensemble median changes in CDD between a 1.5 or 2.0 °C warmer world and the baseline period (1986)(1987)(1988)(1989)(1990)(1991)(1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005). The changes in CDD with warming exhibit a dipole pattern over Eurasia, featuring longer dry spells in the Asian monsoon-Mediterranean region and shorter dry spells in the vast northeastern Asia. The SRC region lies between, thus exhibiting longer dry spells in the southwestern part and shorter ones in the northeastern part. Among the SRC subregions, the reduction in CDD over the regions of Qinghai Province and northern Xinjiang Province is significant. The additional 0.5 °C under the warming of 2 °C increases (decreases) CDD  (4) Changes in CDD are regionally dependent, with longer (shorter) dry spells in the southwestern (northeastern) part. The additional 0.5 °C under warming of 2 °C increases (decreases) the CDD by up to six (five) days in the southwestern (northeastern) part of the SRC region. A uniform increase in extreme heat is seen across the SRC region under the 1.5 °C warming scenario, with the additional 0.5 °C under the 2 °C global warming scenario significantly increasing the number of extreme heat events.
In summary, our comparison of mean and extreme climate changes under 1.5 °C and 2 °C global warming scenarios highlight the impacts that can be avoided by achieving global warming of half a degree lower. If we pursue efforts to limit the temperature increase to 1.5 °C above pre-industrial levels, as compared to holding the increase in the global average temperature to well below 2.0 °C, a further warming of 0.73 °C (0.49-0.94 °C), as well as increased extreme heat events, could be avoided in the SRC region, at a cost of a reduced increase of 2.72% (0.47%-3.82%) in annual precipitation.
trade corridors, power plants, and industrial infrastructure that will share the benefits of more than $1 trillion in economic development. Thus, information on the potential change in climate over the SRC region is of crucial importance to planning future economic activities and hence the successful implementation of the Belt and Road Initiative. In this study, both mean and extreme climate changes projected by the ensemble mean of CMIP5 models are presented. The major findings are summarized in Table 2, but explained in slightly more detail as follows: (1) By the end of the twenty-first century, the multi-model ensemble mean projects a warming of ~1.5, 2.9, 3.6, and 6.0 °C under RCP2.6, 4.5, 6.0, and 8.5, respectively, with respect to the 1986-2005 baseline. These warming magnitudes are around 0.2-0.8 °C greater than the global land average, and 0.5-2.0 °C greater than the global mean. Annual mean precipitation increases consistently across all RCPs. Under RCP8.5, the annual mean precipitation amount increases by ~14% with respect to 1986-2005 over the SRC region by the end of twenty-first century, as compared to ~6.5% for the global land area and ~6.8% for the globe as a whole.
(2) The Millennium Silk Road is projected to witness a 1.5 and 2 °C warming earlier than the global land area and the global average. A regional warming of 1.5 °C (2 °C) over the SRC region is reached before 2020 (2030) under all the emission scenarios. The 2020s and 2030s see a 2 °C warming world, depending on whether RCP8.5 or RCP4.5 is used in the projection. (3) A 1.5 °C warmer world is projected to induce a regional annual mean warming of 1.24 °C over the SRC region relative to the 1986-2005 baseline period, with an interquartile range of 0.92-1.63 °C. The extra 0.5 °C under global warming of 2 °C leads to a further warming by 0.73 °C (0.49-0.94 °C), which is much greater than the global mean level. The annual mean precipitation amount increases by 4.77% (2.31%-8.59%) at the 1.5 °C warming level, with the extra 0.5 °C under warming of 2 °C expected to induce an additional 2.72% (0.47%-3.82%) increase in precipitation. Table 2. projected changes over the src region under global warming of 1.5 °c relative to the 1986-2005 baseline period and the extra 0.5 °c warming relative to 1.5 °c warming using cmip5 models. the numbers in parentheses represent inter-quartile ranges of projected changes.