Effects of spring Arctic sea ice on summer drought in the middle and high latitudes of Asia

Dong Chen , , , Y Go , , Ying Zhng , , , To Wng , ,

a Nansen-Zhu International Research Centre, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China

b Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters/International Joint Laboratory on Climate and Environment Change, Nanjing University of Information Science & Technology, Nanjing, China

c Climate Change Research Center, Chinese Academy of Sciences, Beijing, China

Keywords:Arctic sea ice Summer drought in Asia Snow depth Soil moisture

ABSTRACT Based on data observed from 1979 to 2017, the influence of Arctic sea ice in the previous spring on the first mode of interannual variation in summer drought in the middle and high latitudes of Asia (MHA) is analyzed in this paper, and the possible associated physical mechanism is discussed.The results show that when there is more sea ice near the Svalbard Islands in spring while the sea ice in the Barents–Kara Sea decreases, the drought distribution in the MHA shows a north–south dipole pattern in late summer, and drought weakens in the northern MHA region and strengthens in the southern MHA region.By analyzing the main physical process affecting these changes,the change in sea ice in spring is found to lead to the Polar–Eurasian teleconnection pattern, resulting in more precipitation, thicker snow depths, higher temperatures, and higher soil moisture in the northern MHA region in spring and less precipitation, smaller snow depths, and lower soil moisture in the southern MHA region.Such soil conditions last until summer, affect summer precipitation and temperature conditions through soil moisture–atmosphere feedbacks, and ultimately modulate changes in summer drought in the MHA.

1.Introduction

The region of the middle and high latitudes of Asia (MHA) mainly includes eastern Russia, northern China, Mongolia, and Central Asian countries.The climate conditions in this area are extremely harsh and the human population is sparse, but it is incredibly rich in natural resources.In addition, the China–Mongolia–Russia economic corridor,which has been vigorously developed by many countries in recent years,mainly passes through this region, and the impact of this regional development on the global economy is becoming increasingly important.However, the exploitation and transportation of natural resources as well as the construction and development of economic corridors are closely related to climate change in this region ( Zhou et al., 2020 ;Fan and Li, 2020 ), and extreme climate changes can have very important impacts on these industries.Therefore, conducting research on the characteristics and mechanisms of extreme climate changes, such as changes in drought, in the MHA can not only provide a scientific basis for effectively preventing and reducing the losses caused by extreme climate changes but can also facilitate the better development and exploitation of natural resources in this region.

Since the MHA is adjacent to the Arctic, previous studies conducted in this region have found that the MHA region has a high climate sensitivity and is greatly affected by Arctic sea ice change ( He et al., 2020 ;Li et al., 2020 ).Against the background of global warming, this region is warming with significantly higher temperature amplifications than those measured in other regions ( Allen et al., 2018 ).In terms of precipitation, the northern region of the MHA has experienced increased precipitation over the past few decades, while precipitation in the southern region has decreased ( Seneviratne et al., 2021 ).Under the background of such long-term trends in hydrothermal conditions, an extensive amount of research has focused on the causes of drought events in different parts of the MHA region ( Li et al., 2018 ; Liu et al., 2020 ;Chen et al., 2020 ).Chen et al.(2020) pointed out that interdecadal variations in Arctic sea ice can lead to changes in the Okhotsk high by affecting the westerly trough ridge, ultimately leading to significant interdecadal changes in summer drought conditions in Northeast China.Li et al.(2018) revealed that the springtime reduction in sea ice in the Barents Sea leads to the intensification of drought events in Northeast China by inducing changes in Eurasian snow depth (SD) and soil moisture.Zhu et al.(2019) analyzed sea surface temperatures (SSTs) in the Pacific and Indian Oceans to determine their contribution to the interdecadal enhancement of drought in Northwest China and the associated physical processes.In addition to studies conducted in northern China,some researchers have analyzed the possible causes of historical drought and heatwave events in Central Asian countries and Russia ( Dole et al.,2011 ; Rahmstorf and Coumou, 2011 ; Otto et al., 2012 ; Schubert et al.,2014 ).Otto et al.(2012) revealed that the occurrence of extreme dry heat events in Eurasia is not only related to natural internal variabilities but is also closely related to global warming caused by human activities in recent decades.

However, these studies mainly focused on small-scale drought or unique drought events.Few studies have assessed the overall characteristics of change in drought across the whole MHA region.The external forcing factors affecting interannual changes in drought in the MHA region and the main physical mechanisms associated with these changes are still unclear.As the main water vapor source for precipitation in the MHA comes from westerly transport, there are significant connections among drought events in different regions in this area.It is necessary to perceive the drought characteristics in this region as a whole to provide scientific support for the prediction of future drought events in different countries and areas within this region.Therefore, starting from an analysis of the interannual variation in drought in the MHA region, this paper obtains drought variation characteristics in this region as a whole and the relationships between different regions, analyzes the changes in atmospheric circulation corresponding to the main drought mode, and discusses the physical drought-formation mechanism.

2.Data and methods

Monthly horizontal wind, geopotential height, sea level pressure,air temperature, and specific humidity atmospheric data were obtained from the NCEP (National Centers for Environmental Prediction) reanalysis product and covered the period from 1979–2017; this product is provided by NOAA (the National Oceanic and Atmospheric Administration) and has a horizontal resolution of 2.5° latitude/longitude( Kalnay et al., 1996 ).High-resolution precipitation and near-surface temperature data were derived from Climate Research Unit data (CRU TS4.03), with a resolution of 0.5° × 0.5° ( Harris et al., 2020 ) (obtained from https://crudata.uea.ac.uk/cru/data/hrg/ ).Monthly Arctic seaice concentration (SIC) datasets were obtained from HadISST1 (the Hadley Centre Sea Ice and SST dataset), with a resolution of 1° latitude/longitude ( Rayner et al.2003 ).Monthly SD datasets were obtained from ERA5 (the fifth major global reanalysis produced by ECMWF;https://www.ecmwf.int/en/forecasts/dataset/ecmwf-reanalysis-v5 ).The Self-Calibrating Palmer Drought Severity Index was selected as an index to analyze the interdecadal variation in drought in the MHA( van der Schrier et al., 2013 ).In this study, all data were filtered to remove low-frequency signals over five years, and only interannual variation signals were retained.

A spring sea-ice index was defined as the area-averaged SIC within the domain (78°–83°N, 15°–45°E) minus that within the domain (68°–75°N, 50°–83°E).In this study, the spring season is defined as the mean for the months of March, April, and May (MAM), and summer as the mean for the months of June, July, and August (JJA).

3.Characteristics of the main drought mode

Based on the empirical orthogonal function (EOF) of the interannual variation in drought data collected in the MHA (40°–80°N, 60°–150°E),the first mode of drought in this region has a north–south dipole pattern separated around the 55°N line of latitude, the eastern boundary of the north MHA region is 130°E longitude, northeastern Asian east of 150°E is merged with the south MHA region, and the interpreted variance in the main mode is 15.6%.When drought intensifies in the north MHA region,it weakens in the southern MHA, and vice versa ( Fig.1 (a)).The temporal evolution of the first mode shows that the interannual variability of the first drought mode has strong quasi-biennial oscillation characteristics, which was also confirmed by spectral analysis results (figure not shown), and the interannual variability characteristics also reflect strong interdecadal variability.The interannual variability is stronger before 1995 and after 2005, and very weak from 1995–2005.The strongest variability occurs after 2005 ( Fig.1 (b)).According to previous studies( Chen and Sun, 2015 ; Li et al., 2018 ), the two most important factors affecting drought are changes in precipitation and temperature.Therefore, we analyzed precipitation and temperature changes corresponding to the first mode of the MHA summer drought and found that when the summer MHA drought is weak in the north and strong in the south, summer rainfall is abnormally high in the northern MHA, while the southern MHA has a smaller anomaly ( Fig.1 (c)).In contrast, the temperature in the northern MHA is abnormally low in summer, while the temperature in the southern region is higher.Such hydrothermal conditions are consistent with the drought distribution ( Fig.1 (d)).

4.Possible physical mechanisms

To investigate the possible mechanisms leading to such dipole patterns in precipitation and temperature in the MHA, both of which affect changes in drought in this region, the atmospheric circulation and external forcing factors corresponding to this main drought mode were further analyzed, and the main mode was found to be strongly related to Arctic sea ice in the early spring.When a dipole sea-ice anomaly occurs in the Barents Sea and Kara Sea in the spring, the drought distribution in the MHA reflects a reverse change from north to south later in summer ( Fig.2 (a)).From 1979–2017, the correlation coefficient between the Arctic sea-ice index and the first principal component (PC1) of the MHA drought reached 0.56 and passed the significance test at 0.01 level( Fig.2 (b)).We also calculated the relationships between sea-ice anomalies in key regions and the geopotential height in the Northern Hemisphere, and the results were consistent with the findings of previous studies –namely, that Arctic sea-ice anomalies can cause changes in the Polar–Eurasian teleconnection (POL) pattern by affecting the turbulent heat flux ( He et al., 2018 ; Li et al., 2018 ; Han et al., 2021 ).Corresponding to the change of sea ice in key areas of the Arctic, the geopotential height at 850 hPa and 500 hPa reflects a low-pressure anomaly in the polar and high-latitude regions of Asia and a high-pressure anomaly in northern China and the Mongolian high region (a POL-like pattern)( Fig.2 (c, d)).However, there are also some differences in the geopotential height fields in the lower and middle troposphere.At 850 hPa, the north–south boundary of the high- and low-pressure anomaly center is about 60°N ( Fig 2 (c)), while at 500 hPa the boundary moves about 5°of latitude north ( Fig.2 (d)), which also shows that this is a baroclinically structured system.The vertically integrated water vapor transport (integrated from the surface to 300 hPa) anomaly corresponding to such atmospheric circulation is shown in Fig.3 (a), which will lead to the increase in water vapor transport to the north of 60°N and the decrease of to the south.Since the main water vapor transport pathway in the whole MHA region comes from westerly transport, such abnormal conditions of water vapor transport are conducive to the increase (decrease) of precipitation in the north (south) of the MHA ( Fig.3 (b)).The springtime temperatures in the MHA are generally below zero, and such abnormal precipitation conditions ( Fig.3 (b)) can lead to increased SDs in the northern region of the MHA and decreased SDs in the southern MHA ( Fig.3 (c)).These conditions further intensify the soil moisture in the northern region and lower the soil moisture in the southern region( Fig.3 (d)).Such soil moisture conditions can last until summer.

Fig.1.The (a) first EOF mode of MHA JJA drought during 1979–2017 and (b) its corresponding time series (PC1).(c) Correlation coefficients between PC1 and precipitation.(d) As in (c) but for near-surface temperature.The hatching represents where the correlation coefficients are significant at the 0.1 level based on the Student’s t -test .

Fig.2.(a) Correlation coefficients between PC1 and the SIC in MAM during 1979–2017.The red and blue rectangular areas in (a) represent the selected region for the sea-ice area index (SICI).(b) Time series of the PC1 index (blue dashed line) and SICI index during 1979–2017.(c) Correlation coefficients between the SICI and 850-hPa geopotential height.(d) As in (c) but for the 500-hPa geopotential height.The hatching represents where the correlation coefficients are significant at the 0.1 level based on the Student’s t -test .

Fig.3.(a) Correlation coefficients between the SICI and vertically integrated water vapor transport (vectors, with shading indicating statistical significance at the 0.1 level).(b-d) Correlation coefficients between the SICI and springtime (b) precipitation, (c) snow depth, and (d) soil moisture.The hatching represents where the correlation coefficients are significant at the 0.1 level based on the Student’s t -test.

Fig.4.(a-c) Correlation coefficients between the SICI and summertime (a) precipitation, (b) near-surface temperature, and (c) 850-hPa geopotential height.The hatching represents where the correlation coefficients are significant at the 0.1 level based on the Student’s t -test.(d) Correlation coefficients between the SICI and summertime vertically integrated water vapor transport (vectors, with shading indicating statistical significance at the 0.1 level).

In summer, the continuous melting of snow and the soil moisture changes on the underlying land surface alter the atmospheric circulation, temperature, and precipitation conditions by affecting specific changes in water and heat fluxes.First, the soil moisture in the northern MHA region is high, while that in the southern region is low.Through soil moisture–precipitation feedback effects, soil moisture–temperature feedback effects, and soil moisture–radiation relationships, the summertime moisture content and atmospheric cloud cover change over the southern and northern MHA regions ( Vogel et al., 2018 ).Finally, in the MHA, more precipitation occurs in the northern region and less precipitation occurs in the southern region in summer ( Fig.4 (a)).At the same time, the northern region’s temperatures are low, while the southern region’s are high ( Fig.4 (b)).As a result, the higher precipitation in the northern region leads to smaller evapotranspiration, while the lower precipitation in the southern region leads to higher evapotranspiration; thus, drought weakens in the northern MHA and strengthens in the southern MHA.

In addition, Fig.4 (c) shows the relationship between springtime sea ice in key areas and the 850-hPa geopotential height above the MHA in summer.Similar to the POL pattern observed in spring, the MHA still has a north–south dipole pattern in summer.The water vapor transport (integrated from the surface to 300 hPa) corresponding to such an anomalous geopotential height is shown in Fig.4 (d).Such anomalies superimposed on the climatology lead to increased water vapor transport in the northern MHA and weakened water vapor transport in the southern MHA.This result reflects another cause of the summer drought weakening in the northern MHA and strengthening in the southern MHA.Since spring sea ice is a signal of the previous season and is measured one season ahead of summer drought, when forecasting late summer drought in the MHA, the spring sea ice in the key region can be used as an effective early predictor.

5.Conclusions and discussion

This paper mainly analyzes the characteristics of the first interannual variability mode of summer drought in the MHA and discusses the influence of early spring sea ice on drought in this area as well as the main physical processes associated with this influence.We found that the influence of the early spring Arctic sea ice on the subsequent summer drought takes place mainly through the Eurasian snow from spring to summer.Then, in summer, through the melting of snow, the soil moisture in the MHA area changes, and ultimately affects the subsequent drought through soil moisture–precipitation feedback and soil moisture–temperature feedback.Since the summer precipitation and temperature caused by sea-ice change show opposing north–south changes, the dominant mode of drought in the MHA region also presents a dipole pattern.

In this work, the two processes of soil moisture–precipitation feedback and soil moisture–temperature feedback ( Vogel et al., 2018 ) are the key to the impact of early sea ice on subsequent drought in the MHA region.However, both temperature and precipitation have a very important impact on drought.The specific contribution of soil moisture–temperature feedback and soil moisture–precipitation feedback to drought are not given quantitatively, which needs to be further studied in follow-up work.

Funding

This research was jointly supported by the National Key R&D Program of China [grant number 2017YFE0111800 ] and the National Science Foundation of China [grant numbers 41991281 and 41875110 ].