Heat budget analysis in three typical warm periods simulated by FGOALS-s2

LUAN Yi-Hu， YU Yong-Qingnd ZHENG Wei-Peng

aState Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics， Institute of Atmospheric Physics，Chinese Academy of Sciences， Beijing， China；bClimate Change Research Center， Chinese Academy of Sciences， Beijing， China

Heat budget analysis in three typical warm periods simulated by FGOALS-s2

LUAN Yi-Huaa，b， YU Yong-Qiangaand ZHENG Wei-Penga

aState Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics， Institute of Atmospheric Physics，Chinese Academy of Sciences， Beijing， China；bClimate Change Research Center， Chinese Academy of Sciences， Beijing， China

This study compared basic warming patterns among three typical warm periods — the mid-Holocene （MH）， Medieval Warm Period （MWP）， and the twentieth century warming （20CW） —and carried out a comprehensive heat budget analysis using four experiments simulated by the Flexible Global Ocean—Atmosphere—Land System model， Spectral Version 2 （FGOALS-s2）. The model simulates similar spatial warming patterns in all three warm periods， e.g. stronger warming appears in the high latitudes. However， changes in surface air temperature （SAT） over the tropical regions are different: a significant warming occurs in the 20CW and MWP but a significant cooling in the MH. The heat budget analysis suggested that SAT changes are mainly induced by the heat flux. In the MH， the insolation and positive snow and ice feedback are responsible for the warming in the Southern Ocean but the wind anomalies and decreased downward longwave radiation （DLR） induce the cooling in the tropics. In the 20CW， the decreased shortwave radiation and increased sea surface temperature dependency of evaporation dampen the warming in the tropics. In the MWP， the shortwave radiation induces the Southern Ocean warming， but the DLR and wind anomalies warm the SAT in the tropics. The simulated ocean temperature and ocean heat content anomalies are different in the upper ocean （above 1500 m）， which are mainly induced by the wind stress changes，but similar in the deep ocean in all three warm periods.

ARTICLE HISTORY

Accepted 25 September 2015

Surface air temperature； heat flux； mid-Holocene； Medieval warm period； twentieth

century warming

本文利用 FGOALS-s2 的四组试验，对比了三个典型温暖时期：中全新世时期、中世纪暖期和20世纪增暖期的基本增暖模式，并进行了热收支分析。发现模式模拟的三个暖期的增暖有一些类似的分布，例如：在高纬度特别是南大洋区域都出现了强的增暖。但在热带地区的变化存在差异：中世纪暖期和20世纪增暖期在热带地区存在显著的增暖，但是中全新世时期在热带地区变冷。热收支分析表明大气表层温度的变化主要是由热通量引起的。模式模拟的三个暖期的上层海洋温度和热含量的变化也不相同，主要是由风应力的变化引起的，但深层海洋温度的变化是一致的。

1. Introduction

Against the global warming background， the climate system （e.g. atmospheric and oceanic circulation） has undergone significant changes. To better understand the global warming since the preindustrial （PI） era， it is important to extend our scope of investigation to the paleoclimate. The mid-Holocene （MH， 6 ka BP）， MWP（around AD 900—1300）， and the twentieth century warming （20CW） are three important and typical warm periods in the Quaternary since the start of the Holocene. These three typical warm periods were driven by different external forcings （Solomon et al. 2007； Stocker et al. 2013）. In the MH， the most important external forcing was the seasonal changes of insolation induced by the changes in orbital parameters. The external forcing for the MWP was changes in solar radiance and volcanic eruptions， in which there were fewer volcanic eruptions but greater solar radiation input. And the 20CW is mainly being driven by anthropogenic greenhouse gas （GHG）emissions.

Previous studies have investigated the past climate changes in specific periods such as the MH and MWP relied on proxy data （Marcott et al. 2013； Wang et al. 2002）or climate models （Wang， Wang， and Jiang 2010； Zhou et al. 2011）. Marcott et al. （2013） indicated that the global mean annual temperatures around the MH were about 0.7 °C higher， and extratropical Northern Hemisphere（NH） temperatures were about 1 °C higher， than under PI conditions. Numerical simulations show broad agreement with reconstructions （Braconnot et al. 2007）， albeit with some mismatch in annual and winter temperatures（Jiang et al. 2012； Liu et al. 2014）. The climate of the MWP was warmer than that in the past millennium（Wang et al. 2002）， but the warming magnitude was weaker than that in the twentieth century （Zhou et al. 2011）.

Table 1.Annual mean， boreal summer and boreal winter changes of global mean net radiative flux at the top of the atmosphere （Net），SAT， SST， OHC in the upper 700 m of the ocean and OHC in the whole ocean in the MH， 20CW， and MWP.

A number of studies have shown that the changes in ocean heat content （OHC） played an important role in the changes of Earth's heat balance in the late twentieth century， demonstrating that OHC more reliably represents the response of Earth's energy budget to radiative perturbations than do surface temperatures （Hansen et al. 2005； Levitus et al. 2000， 2001， 2012）. Rosenthal，Linsley， and Oppo （2013） used high-resolution proxy records to show that the changes in OHC were also larger than those in surface temperature during the MH and MWP.

However， the above-mentioned studies only focused on one specific period； understanding the similarities and differences among the characteristics and mechanisms of warming in the aforementioned three warm periods （MH， WMP， 20CW） is very important for projecting future climate change. The Flexible Global Ocean—Atmosphere—Land System model， Spectral Version 2（FGOALS-s2） has carried out most of the Coupled Model Intercomparison Project， Phase 5 （CMIP5） experiments— including simulation of the MH， the last-millennium（LM） climate simulation， and the historical simulation —providing an excellent opportunity to better understand the warming mechanisms in those three periods. Based on these simulations with FGOALS-s2， the objectives of the present study were to compare the spatial patterns of atmosphere and ocean warming in the three warm periods and explore the possible causes using heat budget analysis.

The rest of the paper is organized as follows: Section 2 describes the model data used in the study. Section 3 reports the results from the model's simulations. And finally， a summary and discussion are provided in Section 4.

2. Model data

The datasets used in the present study were from four groups of experiments — the PI simulation， historical simulation， LM simulation， and MH simulation — simulated by FGOALS-s2 （see Bao et al. （2013） for more details）. The simulations have several experiment members， but one member （r1i1p1） in each simulation was used in this study. These experiments are all standard runs based on the CMIP5 and PMIP3 experimental design （Braconnot et al. 2012； Taylor， Stouffer， and Meehl 2012）. To investigate the climate changes in the MWP and 20CW periods， we chose the core time period of the MWP （AD 1150—1199） from the LM simulation and the warmer period of the 20CW（1956—2005） from the historical simulation， respectively，to compute the climatological mean. Correspondingly， we considered the last 50 years from the PI and MH simulations. The analysis of the results made in this paper was based on the changes in the MH （differences between the MH and PI）， 20CW （differences between the 20CW and PI），and MWP （differences between the MWP and LM）.

3. Results

Previous studies have suggested that FGOALS-s2 is able to capture the major characteristics of modern and paleo climate changes， especially the large-scale features of the atmospheric and oceanic circulations （Bao et al. 2013；Zheng and Yu 2013）. Compared to the PI， the major forcing in the MH was the changes in the seasonal cycle of insolation. The MH was characterized by an enhanced seasonality of insolation in the NH and a reduced seasonal cycle of insolation in the Southern Hemisphere（SH） （Berger 1978； Luan et al. 2012）. In boreal summer（June—August）， more net radiative flux at the top of the atmosphere （TOA） was received in the MH than that in the PI （Table 1）， accompanied by a warming over the mid and high latitudes but cooling over the tropical regions（not shown）. In the 20CW period， the net radiative flux at TOA has not changed much， not only in boreal summer but also in boreal winter （December—February） （Table 1）. The warming has mainly arisen from increased anthropogenic GHG emissions. The MWP climate was characterized mainly by natural variability due to changes in solar variability and volcanic eruptions. The global averaged surface air temperature （SAT） anomalies were positive during boreal summer and winter in the MWP， compared with the LM （Table 1）.

Figure 1.Annual mean changes of SAT （contour lines； units: °C） and wind stress （arrows； units: N m-2） in the （a） MH， （b） 20CW， and （c）MWP. Color shading indicates statistically significant SAT changes at the 95% confidence level using the t-test.

The spatial patterns of the annual mean changes in SAT and wind stress in the three warm periods are illustrated in Figure 1. In the MH， FGOALS-s2 simulates warming in Northern Eurasia， the Arctic， Antarctic， and Southern Ocean around 60°S. Significant cold anomalies are found in the tropical regions and middle latitudes （Figure 1a）. In the 20CW， significant warming occurs at the global scale. However， the SAT warming over the tropics is weaker than that in the high latitudes （Figure 1b）. In the MWP，the changes of SAT are smaller than those in the MH and20CW. FGOALS-s2 simulates a weaker global warming in the MWP compared with the LM， especially in the tropical regions （Figure 1c）. It is notable that in all three warm periods the warmer SATs appear in the high latitudes，particularly in the Southern Ocean around 60°S. However，the SAT changes over the tropical regions are different in the three warm periods. FGOALS-s2 simulates significant annual cooling in the MH， but annual warming in the 20CW and MWP.

Figure 2.Changes in regional and annual mean （a， c， e） TOA radiative flux， including net heat flux （Net） and SR， and （b， d， f） surface heat flux， including net heat flux （Net）， SR， DLR， LHF， and SHF， in the （a， b） MH， （c， d） 20CW， and （e， f） MWP. Downward is positive； units: W m-2.

Figure 3.Changes in regional and annual mean LHF， the oceanic response of LHF （LHF_OcF）， and the atmospheric forcing of LHF（LHF_AtF）， in the （a） MH， （b） 20CW， and （c） MWP. Downward is positive； units: W m-2.

To understand the changes in heat budget in various regions， we divided the globe into five bands: 45°S—90°S，15°S—45°S， 15°S—15°N， 15°N—45°N， and 45°N—90°N. Figure 2 shows the changes in the regional and annual mean TOA radiative flux and surface heat flux， including the net heat flux （Net）， shortwave radiation （SR）， downward longwave radiation （DLR）， latent heat flux （LHF）， and sensible heat flux （SHF）. In the MH， the TOA SR increases in the high latitudes and decreases in the middle latitudes and tropical regions （Figure 2a）. Increased insolation in the high latitudes warms the SAT and then weakens the snow and ice albedo， leading to a further increase in absorbing SR at the surface （Figure 2b）. The temperature increases in the high latitudes， induced by the larger amount of solar energy and the positive snow and ice feedback. In the tropical regions （15°S—15°N）， decreased DLR at the surface induces the cooler SAT in the MH and offsets the effectof LHF （Figure 2b）. In the 20CW， the TOA SR increases in the mid and high latitudes but decreases in the tropics（Figure 2c）. At the surface， increased GHG concentrations enhance DLR from 1.5 to 5.5 W m-2in the global region（Figure 2d）， warming the global SAT. Whereas， in the tropical region， the decreased SR offsets the GHG effect （Figure 2d）， damping the SAT warming. In the MWP， the insolation increases almost globally （Figure 2e）， but the surface heat fluxes are smaller than those in the MH and 20CW. In the high latitudes of the SH， the SR at the surface warms the SAT， whereas the surface DLR warms the SAT in the middle latitudes and tropics. The LHF acts to cool the surface air by about -0.3 W m-2（Figure 2f）.

Figure 4.Annual mean changes in the vertical profile of （a—c） air temperature and （d—f） ocean temperature in the （a， d） MH， （b， e） 20CW，and （c， f） MWP. Units: °C.

As shown in Figure 2， the contributions of the LHF to the SAT changes are very important in the three warm periods. We decomposed the LHF into the atmospheric forcing and oceanic response， following Du and Xie （2008）. It is noteworthy that because the LHF over the ocean surface is larger than that over the land surface， the discussion below focuses only on the ocean surface LHF. The atmospheric forcing of LHF （LHF_AtF） is due mostly to atmospheric adjustments in wind speed， surface relative humidity， and the sea minus air surface temperature difference （∆T）. The oceanic response of LHF （LHF_OcF） arises from the sea surface temperature （SST） dependence of evaporation and may approximately act as a Newtonian cooling term. In the MH， the two components of LHF act to slightly dampen the SST warming in the Southern Ocean by about -0.5 W m-2. In the middle latitudes and tropical regions， LHF_OcF acts to warm the ocean， but LHF_AtF acts to cool the ocean（Figure 3a）. In the 20CW， the atmospheric forcing of LHF acts to warm the ocean in the global region； however，LHF_OcF balances the atmospheric forcing （Figure 3b）. In the MWP， the roles of the two components are similar withthose in the 20CW but with smaller magnitude （Figure 3c）. The atmospheric forcing via LHF is mainly due to changes in wind speed in the MH and 20CW （Figure 1）. But the LHF_AtF due to changes in surface relative humidity and∆T may not be negligible in the ocean heat budget （Du and Xie 2008）.

As shown above， the changes in SAT are mainly influenced by the radiative flux and wind anomalies. We further illustrate the vertical structure of the changes in the air and ocean temperatures during the three warm periods in Figure 4. The vertical pattern of the air temperature in the MH exhibits a cooling anomaly in the middle and low latitudes， with the cooling center of about -0.6 °C located at around 200—400 hPa and a warming anomaly in the high latitudes （Figure 4a）. The 20CW and MWP simulations feature similar vertical patterns in the atmosphere， with the warming center located at around 200—400 hPa （Figure 4b and 4c）. Nevertheless， the warming of the MWP is significantly weaker than that of the 20CW （0.2 vs. 1.8 °C）. Zhou et al. （2011） suggested that the mid-tropospheric amplification of temperature is determined by the convective heating. In the ocean， the changes in simulated ocean temperature are different in the upper ocean （above 1500 m）， but similar in the deep ocean， in the three warm periods （Figures 4d—f）， and this is because the upper ocean temperature is mainly induced by the wind stress changes. In the MH， the upper ocean temperature is one of warming in the Southern Ocean but cooling in other regions （Figure 4d）. In the 20CW， warming appears in the upper ocean above 500 m globally （Figure 4e）. There are two cooling centers at around 1000 m in the 20CW （Figure 4e）. In the MWP simulation， there is weak cooling in the tropics but warming in the middle and high latitudes in the upper ocean（Figure 4f）. The annual mean OHC above 700 m decreases in the MH and increases in the 20CW and MWP （Table 1）. The changes in annual mean OHC in the upper 700 m of the ocean are consistent with the SST changes in the three warm periods but with larger magnitude. The temperature anomalies in the deep ocean are warming anomalies，with the maximum located at around 3000 m and mainly induced by the thermohaline circulation. Furthermore，the annual mean changes in the whole ocean OHC of the three warm periods are positive （Table 1）.

4. Summary and discussion

This study compared the spatial patterns of atmospheric and oceanic warming and analyzed the heat budget in three typical warm periods using FGOALS-s2. Because of the different external forcings， the responses of SAT are different in the three warm periods. In the MH， despite being referred to as a warm period， FGOALS-s2 simulates warming in northern Eurasia， the Arctic， Antarctic， and Southern Ocean around 60°S， but significant cooling in the tropics. In the 20CW and MWP， the model captures global-scale warming， with the warming being stronger over the land than over the ocean. Whereas， the SAT changes in the MWP are weaker than those in the 20CW due to the weaker surface radiative forcing.

Changes in SAT are mainly induced by the surface heat flux. In the MH， more net solar energy is received in the high latitudes， inducing the increased surface SR and warming the SAT. And the positive snow and ice feedback enhance the warming. However， in the tropics， the decreased DLR and atmospheric adjustments via LHF induce the SAT cooling. In the 20CW， increased GHG concentrations increase the DLR and warm the global SAT. However， the decreased SR and increased SST dependency of evaporation balances the positive forcing， damping the warming in the middle latitudes and tropics. In the MWP， solar radiation plays an important role in the Southern Ocean warming， but the DLR and atmospheric adjustments via LHF are more important in other regions.

The vertical structures of the changes in the air and ocean temperatures in the three warm periods were also compared. In the atmosphere， a similar pattern appears in the 20CW and MWP， with the warming center located at around 200—400 hPa. However， the warming of the MWP is significantly weaker than that in the 20CW. The air temperature anomalies are different in the MH and exhibit a cooling anomaly in the middle and low latitudes and a warming anomaly in the high latitudes. In the ocean，in addition to the Southern Ocean， the upper ocean temperature （above 1500 m） is one of cooling in the MH but warming in the 20CW and MWP， and this is because the wind stress changes are mainly responsible for the upper ocean temperature changes. There are two cooling centers at around 1000 m in the 20CW. The changes in annual mean upper OHC （700 m） are consistent with the SST changes in the three warm periods， but with larger magnitude. The temperature anomalies in the deep ocean are warming anomalies， with the maximum located at around 3000 m and mainly induced by the thermohaline circulation.

This study suggests that the warming mechanisms and patterns differ among the three warm periods. The SAT changes simulated by FGOALS-s2 are broadly consistent with other CMIP5 models， albeit with some non-significant mismatches in the Southern Ocean in the 20CW （figure not shown）. Because of the uncertainty in the model simulations， a multi-model or model—data comparison would be needed to better understand the mechanisms of global warming. The study also indicates that the OHC in the deep ocean is more sensitive to the external forcing than the SST in these warm periods. With no observed OHC in thedeep ocean， it is difficult to assess the simulation ability of the model. In the future， it is necessary to strengthen the availability of observations and studies of OHC in the deep ocean.

Acknowledgement

We thank the climate modeling groups for producing and making available their model output.

Funding

This study was jointly supported by the National Natural Science Foundation of China ［grant numbers 41406045 and 41376002］； National Basic Research Program of China［grant number 2013CB956204］； ‘Strategic Priority Research Program on Climate Change: Carbon Budget and Relevant Issues' of the Chinese Academy of Sciences ［grant number XDA05110302］.

References

Bao， Q.， P. Lin， T. Zhou， Y. Liu， Y. Yu， G. Wu， B. He， et al. 2013. “The Flexible Global Ocean-atmosphere-land System Model，Spectral Version 2: FGOALS-S2.” Advances in Atmospheric Sciences 30 （3）: 561—576. doi:10.1007/s00376-012-2113-9.

Berger， A. 1978. “Long-term Variations of Daily Insolation and Quaternary Climatic Changes.” Journal of the Atmospheric Sciences 35: 2362—2367.

Braconnot， P.， B. Otto-Bliesner， S. Harrison， S. Joussaume， J. Y. Peterchmitt， A. Abe-Ouchi， M. Crucifix， et al. 2007. “Results of PMIP2 Coupled Simulations of the mid-Holocene and Last Glacial Maximum — Part 1: Experiments and Large-scale Features.” Climate of the Past 3: 261—277. doi:10.5194/cp-3-261-2007.

Braconnot， P.， S. P. Harrison， M. Kageyama， P. J. Bartlein， V. Masson-Delmotte， A. Abe-Ouchi， B. Otto-Bliesner， and Y. Zhao. 2012. “Evaluation of Climate Models Using Palaeoclimatic Data.” Nature Climate Change 2: 417—424. doi:10.1038/ nclimate1456.

Du， Y.， and S. P. Xie. 2008. “Role of Atmospheric Adjustments in the Tropical Indian Ocean Warming during the 20th Century in Climate Models.” Geophysical Research Letters 35: L08712. doi:10.1029/2008GL033631.

Hansen， J.， L. Nazarenko， R. Ruedy， M. Sato， J. Willis， A. Del-Genio，D. Koch， et al. 2005. “Earth's Energy Imbalance Confirmation and Implications.” Science 308: 1431—1435.

Jiang， D.， X. Lang， Z. Tian， and T. Wang 2012. Considerable Model-data Mismatch in Temperature over China during the mid-Holocene: Results of PMIP Simulations. Journal of Climate， 25:4135—4153. doi:10.1175/JCLI-D-11-00231.1.

Levitus， S.， J. I. Antonov， T. P. Boyer， and C. Stephens. 2000.“Warming of the World Ocean.” Science 287: 2225—2229.

Levitus， S.， J. I. Antonov， T. P. Boyer， O. K. Baranova， H. E. Garcia，R. A. Locarnini， A. V. Mishonov， et al. 2012. “World Ocean Heat Content and Thermosteric Sea Level Change （0—2000 m）， 1955—2010.” Geophysical Research Letter 39: L10603. doi:10.1029/2012GL051106.

Levitus， S.， J. I. Antonov， J. Wang， T. L. Delworth， K. W. Dixon，and A. J. Broccoli. 2001. “Anthropogenic Warming of Earth's Climate System.” Science 292: 267—270.

Liu， Z.， J. Zhu， Y. Rosenthal， X. Zhang， B. L. Otto-Bliesner， A. Timmermann， R. S. Smith， et al. 2014. “The Holocene Temperature Conundrum.” Proceedings of the National Academy of Sciences. E3501—E3505. doi:10.1073/ pnas.1407229111.

Luan， Y. H.， P. Braconnot， Y. Yu， W. Zheng， and O. Marti. 2012. “Early and mid-Holocene Climate in the Tropical Pacific: Seasonal Cycle and Interannual Variability Induced by Insolation Changes.” Climate of the Past 8: 1093—1108.

Marcott， S. A.， J. D. Shakun， P. U. Clark， and A. C. Mix. 2013. “A Reconstruction of Regional and Global Temperature for the past 11，300 Years.” Science 339: 1198—1201.

Rosenthal， Y.， B. K. Linsley， and D. W. Oppo. 2013. “Pacific Ocean Heat Content during the past 10，000 Years.” Science 342: 617—621. doi:10.1126/science.1240837.

Solomon， S.， D. Qin， M. Manning， Z. Chen， M. Marquis， K. B. Averyt， M. Tignor， and H. L. Miller （eds.）. 2007. IPCC， Climate Change 2007: The Physical Sciences Basis. Cambridge: Cambridge University Press.

Stocker， T. F.， D. Qin， G. K. Plattner， M. Tignor， S. K. Allen， J. Boschung， A. Nauels， Y. Xia， V. Bex， and P. M. Midgley（eds.）. 2013. IPCC， Climate Change 2013: The Physical Sciences Basis. Cambridge: Cambridge University Press.

Taylor， K. E.， R. J. Stouffer， and G. A. Meehl. 2012. “An Overview of CMIP5 and the Experiment Design.” Bulletin of the American Meteorological Society 93: 485—498.

Wang， S.， Z. Xie， J. Cai， J. Zhu， and D. Gong. 2002. “Global Mean Temperature Changes during the Last Millennium.” Progress in Natural Science 12 （12）: 920—924.

Wang， T.， H. J. Wang， and D. B. Jiang. 2010. “Mid-Holocene East Asian Summer Climate as Simulated by the PMIP2 Models.”Palaeogeography Palaeocology 288: 93—102.

Zheng， W.， and Y. Yu. 2013. “Paleoclimate Simulations of the mid-Holocene and Last Glacial Maximum by FGOALS.” Advances in Atmospheric Sciences 30 （3）: 684—698.

Zhou， T.， B. Li， W. Man， L. Zhang， and J. Zhang. 2011. “A Comparison of the Medieval Warm Period， Little Ice Age and 20th Century Warming Simulated by the FGOALS Climate System Model.”Chinese Science Bulletin 56 （28—29）: 3028—3041.

30 June 2015

CONTACT LUAN Yi-Hua luanyh@mail.iap.ac.cn

© 2016 The Author（s）. Published by Taylor & Francis.

This is an Open Access article distributed under the terms of the Creative Commons Attribution License （http://creativecommons.org/licenses/by/4.0/）， which permits unrestricted use，distribution， and reproduction in any medium， provided the original work is properly cited.