Seasonal persistence of soil moisture anomalies related to freeze–thaw over the Tibetan Plateau and prediction signal of summer precipitation in eastern China
- 105 Downloads
Soil moisture can be an effective climate prediction signal due to its long memory. This study investigated seasonal persistence of soil moisture anomalies from the preceding autumn to spring dominated by the soil freeze–thaw (FT) process over the Tibetan Plateau (TP), and their relationship with summer precipitation in eastern China. Results demonstrated that soil moisture anomalies from the preceding autumn can persist until spring by water storage effect of the soil FT process. Soil moisture in the TP during the preceding autumn and winter had similar climatic effects as spring soil moisture. Positive soil moisture anomalies in the eastern TP during the spring led to less summer precipitation in south China and the Yellow River basin, and more summer precipitation in the Yangtze River basin and northeast China. A possible mechanism for this was that wetter soil moisture anomalies from the preceding autumn were stored in the soil by soil freezing, and were released with soil thawing in the spring, inducing surface diabatic heating anomalies over the TP. These anomalies then persisted into summer and enhanced the TP’s thermal forcing to the subtropical westerlies and affected stationary Rossby wave train propagation in middle latitudes, particularly on the northwest and northeast sides of the TP. This study suggests that most of spring soil moisture anomalies signal contains the preceding two seasons’ soil moisture anomalies information; therefore, summer precipitation predicting signals can be obtained from soil moisture anomalies from the preceding autumn, which could lengthen the seasonal climate prediction period.
KeywordsSoil moisture anomalies Seasonal persistence Freeze–thaw Tibetan Plateau Summer precipitation Subtropical westerlies
Seasonal climate prediction is a challenging issue, in which finding a robust signal connected with precipitation or temperature variability on seasonal scale is crucial. Sea surface temperature anomalies as currently the most significant signal (e.g., the El Niño/La Niña phenomena) widely used for climate prediction (e.g., Ropelewski and Halpert 1987; Wang and Eltahir 1999; Wang et al. 2000; Zhang et al. 1999; Dai and Wigley 2000). Other factors such as soil moisture (e.g., Fischer et al. 2007; Koster et al. 2003, 2004; Seneviratne et al. 2006; Conil et al. 2007; May et al. 2017; Ardilouze et al. 2018) and snow cover/depth (e.g., Blanford 1884; Chen 1997; Wu and Qian 2003; Ding et al. 2010; Wang et al. 2017; Yang et al. 2017; Diro et al. 2017), have received much attentions. Studies suggested that sea surface temperature has a much smaller effect than soil moisture on predicting precipitation over mid-latitude land (Koster and Suarez 1995; Conil et al. 2007; Yang et al. 2016). Soil moisture affects the weather and the climate (e.g., Koster et al. 2004; Diro and Sushama 2017) through its effects on land surface energy balance and the water cycle (Seneviratne et al. 2010). Soil moisture varies slowly relative to atmospheric elements, with soil moisture anomalies persisting for several weeks to 1 month or longer (several months or seasons) in deep soil layers (Koster and Suarez 2001; Dirmeyer et al. 2008; Li et al. 2016). Soil moisture content influences not only current climatic conditions, but also subsequent climate conditions through its memory effect (Beljaars et al. 1996).
The relationship between spring soil moisture and subsequent summer precipitation has attracted much attention. For example, early work by Namias (1952, 1960) demonstrated that spring precipitation and soil moisture can affect summer precipitation in continental interiors. Modeling and observations have demonstrated that springtime soil moisture conditions can aid in drought or flood year prediction (e.g., Yeh et al. 1984; Findell and Eltahir 1997; Koster et al. 2003; Grimm et al. 2007; Meng and Quiring 2010). Studies have suggested that spring soil moisture anomalies in eastern China (EC) can affect summer monsoon circulation anomalies, leading to local summer precipitation anomalies (e.g., Zuo and Zhang 2007; Zhang and Zuo 2011; Meng et al. 2014). Summer precipitation over EC is also closely related to spring soil moisture anomalies on the Tibetan Plateau (TP; Chow et al. 2008; Wang et al. 2009; Li and Wang 2016). These studies have suggested that soil moisture during the spring could be a signal for summer climate prediction (e.g., Koster et al. 2003; Ma et al. 2000; Guo et al. 2007). For instance, assimilating spring soil moisture can improve predictions of summer precipitation (Walker and Houser 2001; Wang and Cui 2018). Nevertheless, seeking a signal over a longer time scale is still a prime climate prediction goal (Vautard et al. 2007; Quesada et al. 2012). Additionally, exploring factors in effect before spring could provide a better signal for boreal summer climate predictions.
Studies have shown that the FT process significantly affects land surface water and energy budgets though phase changes of soil water (Wang et al. 2003; Guo et al. 2011; Wang and Yang 2018). The FT process over the TP shows a significant connection with atmospheric circulation anomalies in East Asia (Wang et al. 2003, 2008a, b). Additionally, spring soil moisture anomalies induced by thawing of frozen soils and melting of snow affect summer precipitation in EC (Wang et al. 2003, 2017; Chow et al. 2008). Spring serves as a transition from winter to summer, and because the atmosphere is mainly influenced by land and ocean diabatic heating, seasonal climate prediction has a well-known spring disorder. Studies regarding the effects of spring soil moisture on subsequent summer precipitation (e.g., Namias 1952, 1960; Findell and Eltahir 1997; Zhang and Zuo 2011; Meng et al. 2014, Yang et al. 2016, Wang and Cui 2018) have focused on spring diabatic heating. Nevertheless, soil moisture anomalies prior to spring have been suggested to also affect summer climate. For example, heat in European summers can be induced by wet or dry conditions during the winter and spring (Vautard et al. 2007; Quesada et al. 2012). In addition to being a transitional season, spring is also when frozen soils thaw and snow melts over the TP, which has a large distribution of frozen ground. Yang and Wang (2018) have suggested that the soil FT process play a water storage role that can lengthen soil moisture memory and affect spring soil moisture content. This implies that spring soil moisture anomalies might contain information on soil moisture anomalies from the preceding winter or autumn.
In this study, we investigate the relationship between soil moisture anomalies over the TP in seasons before spring (autumn and winter in the previous year) and subsequent summer precipitation in EC and explore possible physical mechanisms by which soil moisture anomalies persisting from preceding seasons affect summer precipitation in EC. Elucidating these aspects would benefit seasonal climate prediction and deepen understanding of FT’s role in land–atmosphere interactions.
The remainder of this paper is organized as follows. In Sect. 2 we introduce data and methodology. In Sect. 3 we analyze the connection between spatiotemporal variations in soil moisture over the TP and summer precipitation in EC. In Sect. 4 we verify statistical results by performing numerical experiments and explore possible mechanisms linked to the impacts of the persistence of TP soil moisture anomalies on summer precipitation in EC. Lastly, we provide a discussion and main conclusions in Sect. 6.
2 Data and methodology
In this study, the daily and monthly soil moisture data for period 1979–2010 were obtained from the Global Land Data Assimilation System (GLDAS) Version 2, generated by the Noah land surface model, with a 1° × 1° spatial resolution. The GLDAS generates satellite- and ground-based observational data using advanced land surface modeling and a data assimilation technique (Rodell et al. 2004). Previous studies have used GLDAS soil moisture data and have verified the data across the TP (e.g., Chen et al. 2013; Bi et al. 2016). Although GLDAS soil moisture data have errors and uncertainties over the TP, the data still possess certain representativeness, even over complex terrain (Bao et al. 2017). The GLDAS soil moisture dataset provides four depth layers: 0–0.1 m, 0.1–0.4 m, 0.4–1.0 m, and 1.0–2.0 m. In this study we focused on soil moisture variations between 0 and 1.0 m.
To analyze connections between soil moisture in the spring and preceding seasons, we further subdivided the duration of the data spanning from September in the previous year to May into three periods: the preceding autumn [September–October–November (PSON)], the preceding winter [December–January–February (PDJF) and spring (MAM)]. The TP covers a region bounded by coordinates 25°–40°N and 75°–105°E and has an elevation greater than 2000 m.
Daily precipitation data for 1979–2010 were collected from 756 observation stations of the National Climate Center of the CMA and we obtained monthly precipitation from the average of daily data.
We employed the EOF to investigate spatiotemporal characteristics of soil moisture anomalies in different seasons.
To validate the statistical results, a series of experiments were performed using the Community Earth System Model Version 1.2.0 (CESM 1.2.0), which was developed by the U.S. National Center for Atmospheric Research and implemented with fully dynamic atmosphere, land, ocean, sea ice, and river runoff components. The atmosphere component is the Community Atmosphere Model Version 4, and the land component is the Community Land Model Version 4 (CLM4). CESM has been widely used in other studies (e.g., Hurrell et al. 2013) and can effectively simulate general circulation and climate characteristics over the TP and surrounding regions (Wang et al. 2017). Studies have also suggested that CLM4 has good performance for land surface processes (Lawrence et al. 2011) and can well describe soil moisture characteristics in FT processes (Yang et al. 2018).
3 Spatiotemporal evolutions of soil moisture anomalies over the TP and relationship with summer precipitation anomalies in EC
Persistence of soil moisture anomalies has received much attention in recent years, with most studies focusing on seasonal-scale or sub-seasonal-scale soil moisture anomalies (e.g., Koster and Suarez 2001; Seneviratne et al. 2006). Considering the soil FT process’s water storage effect (Yang and Wang 2018), frozen ground in cold regions can produce a soil moisture anomaly that persists longer. Soil moisture anomaly persistence can be quantified by soil moisture memory, which is defined as the average time required for lagged autocorrelation of soil moisture dropping below the 99% confidence level (Dirmeyer et al. 2008). This lag is backward in time and the memory implies how far in the past the anomaly arose. For example, a 120-day memory for March means that anomalies in November persisted into March.
4 Validation of persistent climatic effects of soil moisture anomalies using numerical experiments
4.1 Experiment design
The CTL experiment was performed without modifying soil moisture calculated in the model.
- (2)The SE_MAM experiment was performed with increased soil moisture in the eastern TP (30°–37.5°N, 87°–102°E, rectangular box region in Fig. 5a) during MAM.
The SE_PDJF experiment was performed with increased soil moisture in the eastern TP during PDJF.
The SE_PSON experiment was performed with increased soil moisture in the eastern TP during PSON.
The North rule (North et al. 1982) states that only the maximum value of LVs has meaning. Soil moisture anomalies taken in the SEs were based on large interannual soil moisture anomaly regions (with maximum LVs). For example, the soil moisture was increased in the eastern part of the TP in SE_MAM (Fig. 5a), which corresponds to the maximum LVs region in the MAM soil moisture EOF2 spatial pattern (Fig. 3a). Considering that soil moisture’s maximum standard deviation could reach 0.05 mm3/mm3, and that the average soil moisture climatology was approximately 0.25 mm3/mm3 over the TP, the soil moisture increased by 20% on the monthly scale in the SEs, soil moisture in the 0–1.0 m layer was modified.
CESM was run with fully coupled atmosphere, land, sea ice, and ocean components (the component set is B_2000). Initializations were arbitrary. Each experiment was integrated for 20 model years without re-initialization each year, with the first 10 years for model spin-up and the last 10 years’ results used for the analyses. The last 10 years of simulations represented a 10-member ensemble.
Figure 5 shows increased soil moisture in eastern region of the TP during three seasons and the corresponding differences of the subsequent summer precipitation in EC between CTL and the SEs. When MAM soil moisture in eastern TP was increased (SE_MAM), simulated summer precipitation in south China and the Yellow River basin decreased and simulated summer precipitation in the Yangtze River basin and northeast China increased (Fig. 5b). The distribution of precipitation anomalies was consistent with the statistical results (Fig. 4a), although the positive precipitation anomalies in Inner Mongolia were not well reproduced. When PDJF (SE_PDJF) and PSON (SE_PSON) soil moisture were increased in the eastern part of the TP, simulated summer precipitation anomaly patterns (Fig. 5c, d) were similar to the SE_MAM simulations. These results illustrated that preceding season (i.e., PDJF and PSON) soil moisture can significantly affect subsequent summer precipitation in EC.
5 Possible mechanisms for the relationship between seasonal persistent soil moisture anomalies over the TP and summer precipitation in EC
The atmosphere dynamic process over and around the TP is related to thermal regimes at its two sides. In SE_MAM, air temperatures at the north side of the TP were much colder during the summer due to the TP cold anomalies, which led to an increase in the meridional temperature gradient between the TP and air at its north side (Fig. 8b). Based on thermal wind law, zonal wind had positive anomalies and westerly wind flow on the north side of the TP is accelerated (Fig. 8b, c), and westerlies at the downstream exit region from northeast China to Japan would also be enhanced. In experiments SE_PDJF (Fig. 8e, f) and SE_PSON (Fig. 8h, i), the overall patterns of general circulation anomalies over East Asia caused by soil moisture anomalies in preceding winter and autumn are generally similar to the SE_MAM, except for some differences in details of the pattern of zonal wind anomalies.
Figures 8 and 9 show that general circulation and Rossby wave propagating anomalies patterns for the SEs were similar, which implies that no matter what soil moisture anomalies are present during preceding seasons, these through spring soil moisture anomalies affect general circulation. The experiments results illustrated that soil moisture anomalies in seasons preceding spring can serve as a signal for predicting summer precipitation in EC.
6 Discussion and conclusions
How to extract signals of external forcing from the land surface to improve seasonal climate prediction is an open issue. In this study, we analyzed seasonal persistence of soil moisture anomalies from the preceding autumn to subsequent seasons over the TP using both observations and model experiments and investigated the relationship between soil moisture anomalies from the two preceding seasons and summer precipitation in EC. We found that soil moisture anomaly patterns during the two preceding seasons (PDJF and PSON) were similar to pattern during the spring (MAM). In other words, spring soil moisture anomalies were inherited from preceding seasons. Soil moisture anomalies during MAM, PDJF and PSON over the TP have similar climate impacts, and the relationship between MAM soil moisture anomalies over the TP and summer precipitation anomalies in EC can retrospect to PSON and PDJF. When MAM soil moisture has wetter anomalies in the eastern TP, subsequent summer precipitation in south China and the Yellow River basin was deficient and excessive in the Yangtze River basin and northeast China. This was coincident with the effects of the soil FT process on precipitation (Wang et al. 2003). In the SEs, increased soil moisture in the eastern TP during the preceding autumn, preceding winter, and spring, reproduced similar summer precipitation anomaly patterns, illustrating that like spring soil moisture, soil moisture anomalies in preceding seasons over the TP have a significant relationship with subsequent summer precipitation in EC.
The possible mechanisms for the persistence of soil moisture anomalies from the preceding autumn to spring over the TP and its influence on summer precipitation in EC were investigated. A key linkage was that preceding autumn soil moisture anomalies persisted into MAM through a storage effect from the soil FT process (Yang and Wang 2018). Namely, soil moisture anomaly signals from the preceding autumn and winter were stored by the soil FT process that occurred over the TP and then released in the following spring. With positive PSON soil moisture anomalies persisting and appearing in spring, spring soil moisture anomalies changed surface diabatic heating and affected thermal forcing of the TP to the atmosphere during the summer. Additionally, the meridional temperature gradient increased on the north side of the TP, favoring the acceleration of westerlies at the TP’s north. Meanwhile, at middle latitudes, meridional winds and stationary Rossby wave train propagation weakened in the upper troposphere, thus leading to summer precipitation anomalies in EC.
Climate prediction practices and operation have demonstrated that seasonal climate prediction mainly relies on ocean information (for instance, the El Niño/La Niña phenomenon), which influences global climate variability on seasonal to interannual time scales (e.g., Barnston et al. 1997; Mokhov et al. 2011). In addition to the oceanic “signal,” soil moisture is a slowly varying factor compared to the atmosphere itself and can thus influence the atmosphere for several months or seasons (Koster et al. 2004). Spring soil moisture is usually applied to summer climate prediction as a continental signal from one leading season. This study suggests that soil moisture anomaly signals from the preceding autumn and winter can be transferred by the soil FT process to the spring. In other words, soil moisture anomalies in the spring contain signals from at least two preceding seasons, which provides a new clue for advancing seasonal climate prediction.
Remarkable spatial differences in soil moisture anomalies are due to different soil hydrothermal regimes. Previous work has shown that regions with prominent interannual variability of spring soil moisture locate in the mid-latitude transition zone where seasonal frozen ground is common and soil FT cycles commonly occur (Yang et al. 2016). Results of this study also imply that the role of soil FT process in land–atmosphere interaction might be important for climate prediction during Northern Hemisphere summers.
This work was supported by the National Natural Science Foundation of China (91837205, 41661144017, 41801015 and 41805032). We would like to acknowledge the Meteorology Information Center of the Chinese Meteorology Administration (CMA) for providing observational data.
- Chen LT (1997) The influence of winter–spring snow cover over Tibetan Plateau on atmospheric circulation and flood season precipitation in south China. Middle and long-term water temperature forecast corpus, vol 1. China Waterpower Press, Beijing, pp 194–195 (in Chinese) Google Scholar
- Ding F, Sun ZB, Liu JL (2010) Correlations of snow depth over the Tibetan Plateau in winter and spring with precipitation in summer over China. Trans Atmos Sci 32:783–791Google Scholar
- Guo W, Ma Z, Wang H (2007) Soil moisture—an important factor of seasonal precipitation prediction and its application. Clim Environ Res 12:20–28Google Scholar
- Hoskins B, Pearce RP (1983) Large-scale dynamical processes in the atmosphere. Academic Press, LondonGoogle Scholar
- Hurrell JW, Holland MM, Gent PR, Ghan S, Kay JE, Kushner PJ, Lamarque JF, Large WG, Lawrence D, Lindsay K, Lipscomb WH, Long MC, Mahowald N, Marsh DR, Neale RB, Rasch P, Vavrus S, Vertenstein M, Bader D, Collins WD, Hack JJ, Kiehl J, Marshall S (2013) The community earth system model: a framework for collaborative research. Bull Am Meteorol Soc 94:1339–1360CrossRefGoogle Scholar
- Lawrence DM, Oleson KW, Flanner MG et al (2011) Parameterization improvements and functional and structural advances in version 4 of the community land model. J Adv Model Earth Syst 3:M03001Google Scholar
- Li DX, Wang CH (2016) The relation between soil moisture over the Tibetan Plateau in spring and summer precipitation in the eastern China. J Glaciol Geocryol 38:89–99 (in Chinese) Google Scholar
- Li RL, Bao HY, Li KC, Wang CH (2016) The memory and climate effects of global soil moisture. J Glaciol Geocryol 38:1470–1481Google Scholar
- Ma ZG, Wei HL, Fu CB (2000) Relationship between regional soil moisture variation and climate variability over East China. Acta Meteorol Sin 58:278–287 (in Chinese) Google Scholar
- Namias J (1960) Factors in the initiation, perpetuation and termination of drought. Int Assoc Sci Hydrol Publ 51:81–94Google Scholar
- Oleson KW, Lawrence DM, Bonan GB et al (2013) Technical description of version 4.5 of the Community Land Model (CLM) (NCAR Tech. Note NCAR/TN-5031STR). National Center for Atmospheric Research, CO, Boulder, p 420Google Scholar
- Wang CH, Cheng GD, Deng AJ, Dong WJ (2008b) Numerical simulation on climate effects of freezing-thawing processes using CCM3. Sci Cold Arid Reg 1:68–79Google Scholar
- Wang R, Li W, Liu X, Wang L (2009) Simulation of the impacts of spring soil moisture over the Tibetan Plateau on summer precipitation in China. Plateau Meteorol 28:1233–1241Google Scholar
- Wang J, Qi L, He J, Wu Z (2016) Relationship between spring soil moisture in the Tibetan Plateau and summer precipitation in the Yangtze river basin and its possible mechanism. Chin J Geophys 59:3988–3995Google Scholar
- Wei ZG, Huang RH, Dong WJ (2003) Interannual and interdecadal variations of air temperature and precipitation over the Tibetan Plateau. Chin J Atmos Sci 27:157–170Google Scholar
- Yang K, Hu TT, Wang CH (2017) A numerical study on the relationship between the spring–winter snow cover anomalies over the northern and southern Tibetan Plateau and summer precipitation in East China. Chin J Atmos Sci 41:345–356. https://doi.org/10.3878/j.issn.1006-9895.1604.16119 (in Chinese) Google Scholar
- Ye DZ, Gao YX (1979) Meteorology of the Qinghai-Xizang Plateau. Chinese Science Press, BeijingGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.