Mass Variations in the Siberian Permafrost Region Based on New GRACE Results and Auxiliary Modeling
- 431 Downloads
GRACE (Gravity Recovery And Climate Experiment) determines the integral mass variations in the Earth system with a high spatial-temporal resolution. These mass variations should be adequately separated for better understanding of the individual signal contributions. In Siberia, the temporal mass variations are related to hydrological processes including thawing of permafrost layers. Permafrost layers with different thickness cover about 80% of Siberia. These frozen sheets play an important role for sea level rise and the global hydrological water cycle. In this study, the integral mass variations in Siberia are precisely estimated based on the new release of GRACE (RL05a) from GeoForschungsZentrum (GFZ) in Potsdam. In addition, various hydrological contributions (lake level variation, river run-off, etc.) can be estimated from different models and specific data. Here, mass variations in the Siberian permafrost region based on GRACE results and different hydrological models/data [i.e., GLDAS (Global Land Data Assimilation System) and GPCP (Global Precipitation Climatology Project)] are jointly investigated.
KeywordsGRACE Mass variation Hydrological model GLDAS GPCP Permafrost thawing
Since 2002, the GRACE satellite mission is in a near polar orbit for recovering the static and time-variable parts of the Earth’s gravity field (Tapley et al. 2004). The time-variable part of the Earth’s gravity field is related to mass re-distribution in the Earth system. GRACE monitors the sum of all mass variations occurring in the Earth’s system. It cannot directly separate the different sources and effects. Therefore, an efficient separation technique and very precise background models have to be used for understanding mass variations in different regions over the globe. The provided time series of monthly gravity field solutions in the analysis centres (ACs) are already corrected for variations caused by atmosphere and ocean (Flechtner and Dobslaw 2014). Based on time-variable gravity field coefficients from GRACE, a number of studies to estimate secular trends and periodical changes for different regions have been successfully performed, such as mass balance in Greenland (Velicogna and Wahr 2013), hydrological mass variations in the Amazon basin (Werth et al. 2009), mass losses in Antarctica (Sasgen et al. 2013), mass variations in Siberia, Russia (Steffen et al. 2012; Vey et al. 2013; Chao et al. 2011), hydrological storage changes over Siberia (Frappart et al. 2006, 2011; Muskett and Romanovsky 2009), water depletion in India and China (Tiwari et al. 2014; Rodell et al. 2009; Zhong et al. 2009), terrestrial water budget in the Eurasian pan-Arctic (Landerer et al. 2010), hydrological mass variation in Australia (Seoane et al. 2013), global large-scale groundwater variations from satellite gravimetry and hydrological models (Jin and Feng 2013) and contribution of terrestrial water derived by GRACE in polar motion (Jin et al. 2012). A secular trend in terms of Equivalent Water Thickness (EWT) has been estimated for all those regions. Estimated periodical terms include semi-annual, annual and some further periods.
By increasing the time-span of GRACE observations, mass variation signals including secular and (long-) periodical contributions in different regions can be precisely estimated. One of the challenging regions, which is not sufficiently investigated so far, is the permafrost region in Siberia, Russia. During the past decades, the permafrost regime in the Siberian region has experienced significant changes in terms of permafrost thawing due to climate warming. As an example, Yang et al. (2013) found that the average fresh water discharge from the six largest Eurasian rivers to the Arctic ocean has increased by 7%. Lena, Yenisei and Ob are the three largest rivers in the Siberian permafrost region that flow into the Arctic ocean from south to north. Parts of the signal due to permafrost thawing can be detected based on the level (discharge) changes of those rivers. As the permafrost region covers almost 80% of Siberia with a maximum thickness of 1 km in Yakutia (central Siberia), it is expected that permafrost thawing due to climate change plays a key role in mass (water) balance in this region (Steffen et al. 2012). In addition, the permafrost has caught an extreme capacity of carbon-dioxide (CO2) and nitrogen (N) that, if released, will play an important role for the Earth’s ecosystem in the near future (Treat et al. 2014). Therefore, surface and sub-surface mass variations in Siberia based on the gravitational approach from GRACE monthly solutions are determined, using GRACE data covering almost 12 years. In addition, the GLDAS hydrological model (Rodell et al. 2004) has been used to obtain the soil-moisture contribution, and GPCP precipitation data for estimating precipitation trends in the target region.
2 GRACE Data and Analysis
2.1 Processing Procedure
The GRACE L1b data are officially processed by different analysis centres (ACs), e.g. GFZ, JPL and CSR, as well as by other institutes such as ITG (University of Bonn, Germany), AIUB (University of Bern, Switzerland) and TU Delft (Delft University of Technology, The Netherlands) to produce GRACE L2 data in terms of static and time-variable spherical harmonic coefficients. These processing centres apply different strategies and background models.
GSM data: GRACE Level 2 monthly time-variable gravity field products are provided in terms of fully normalized geo-potential spherical harmonic coefficients. We used GFZ solutions from January 2003 to December 2013 (122 months).
Zonal term replacement: The zonal term c2, 0 of the monthly gravity field coefficients is replaced by the solution from Satellite Laser Range observations (Cheng et al. 2013).
Filtering: The monthly gravity field products are affected by correlated noise at higher frequencies. The correlated noise shows up in terms of striping effects in the mass variation results (e.g. gravity anomalies), when represented in a geographical plot. The striping errors are caused by the orbit design of GRACE-type configurations, incomplete reduction of non-tidal high frequencies of mass variations and limitation in the processing of the Earth gravity field. Therefore, an effective filter has to be applied to separate signal from noise. The normalized spherical harmonic coefficients can be smoothed in many different ways (Jekeli 1981; Swenson and Wahr 2006; Kusche 2007; Klees et al. 2008; Davis et al. 2008) of which we have tested several for Siberia.
Mass variation products: For our purpose, the mass variations in the permafrost region are estimated in terms of EWT. Therefore, the filtered spherical harmonic coefficients from the previous step are used to determine the global TWSC based on an approach presented by Wahr et al. (1998).
Glacial Isostatic Adjustment (GIA) corrections: Vey et al. (2013) showed that the ice history over Siberia does not reveal large-scale ice sheets during the last glacial maximum (LGM). In addition, by applying the two ice models RSES and ICE-5G with different lithospheric thicknesses, no significant GIA effects were found in central Siberia (Velichko et al. 2011; Vey et al. 2013).
It should be mentioned that depending on ROI and SOI, appropriate post-processing techniques including filtering have to be used to separate the signals. From the several filter techniques published and used in the last years, the performance of the 1D isotropic Gaussian (Jekeli 1981), the 2D Fan (Zhang et al. 2009) and DDK filters (Kusche 2007) are tested for use in the permafrost region. In addition to these common filters, a de-striping filter has been applied to the spherical harmonics to minimize the effect of errors due to the north-south stripes in GRACE monthly solutions (Swenson and Wahr 2006).
3 GRACE Derived Mass Variations
Statistical values of secular trend estimation for different filters using GFZ-RL05a data in Siberia covering the period of 2003–2013
Gaussian filter (350 km), de-striping + c2, 0 replaced
Fan filter (350 km), de-striping + c2, 0 replaced
DDK3 + c2, 0 replaced
4 Hydrological Mass Variations
Auxiliary hydrological models, e.g. GLDAS (Rodell et al. 2009) and precipitation data from GPCP centres (Adler et al. 2003) are used to compare those estimated GRACE secular trends. GLDAS provides hydrological parameters with a temporal resolution of 1 month and a spatial resolution of 1∘× 1∘ for all land regions north of 60∘ South. GLDAS is generated by an optimal combination of all land (surface) data; e.g. soil-moisture in different depths or snow cover, and constrained by different satellite data. But permafrost contributions such as continuously frozen ground or thawing and freezing processes at the surface are not considered in GLDAS (Rodell et al. 2004). Hydro-climatic changes of permafrost layers in Siberia are complex and also include not stationary pattern changes (Milly et al. 2008). In this study, to be consistent with the procedure of GRACE data analysis, the same processing strategy and filter-techniques were applied to the GLDAS model. We then suppose that the remaining mass variation signals can be considered as permafrost layer changes.
The differences between GRACE and GLDAS time series might be explained by permafrost thawing effects that were not modeled in GLDAS. The secular trends of 1. 4 cm/a and 0. 7 cm/a during the periods of 2003–2008 and 2008–2013 are estimated for the selected station. As GLDAS does not include permafrost contributions, these results may point to permafrost changes in the Lena basin. Therefore, the estimated positive secular trends for the GRACE minus GLDAS results during the period of 2003–2013 may show that permafrost thawing is proceeding in the target region. But permafrost thawing dynamics and geophysical processes behind it are very complex. By increasing the time span and accuracy of GRACE observations (e.g. with the launch of GRACE follow on in 2017) as well as the availability of more precise hydrological models, permafrost thawing processes in Siberia might be better determined and physically interpreted in near future.
5 Conclusions and Outlook
Our investigations of mass variations in Siberia from GRACE and hydrological models mainly focused on the permafrost regime. It should be mentioned that the lower degree of the spherical harmonics (i.e. zonal term c2, 0) has a significant impact on the mass variation estimations. Using different filter techniques (isotropic and non-isotropic) gives slightly different mass variation trends in Siberia. In addition, the 2D Fan-filter with radius 350 km and after replacing of c2, 0 from SLR solutions seems to be an optimal filter for Siberia. Vey et al. (2013) found that 30–60% of total mass variations in the permafrost region of Siberia can be related to surface water storage changes. Thus combined with our recent results, permafrost thawing can reach up to 1. 3 cm/a EWT for the period of 2003–2008. Vey et al. (2013) also studied the potential benefit of combining GRACE gravimetry, satellite altimetry and satellite imagery data. Thus, future studies of temporal mass variations should include data from further space-borne missions, e.g. satellite altimetry and satellite imagery, to constrain hydrological mass variations in the Siberian permafrost region. The separation of GRACE mass variations can be improved by assimilation of lake surface extent measurements and height variations from satellite altimetry (e.g. Jason-2) and hyper-spectral satellite (e.g. Landsat) data with different temporal-spatial resolutions. Based on these three techniques, the real (sub)-surface mass variation pattern in the Siberia might be better understood.
We would like to thank the GeoForschungsZentrum (GFZ) in Potsdam and the German Space Operations Center (GSOC) of the German Aerospace Center (DLR) for providing continuously and nearly 100% of the raw telemetry and L2 data of the twin GRACE satellites.
- Davis JL, Tamisiea ME, Elsegui P, Mitrovica JX, Hill EM (2008) A statistical filtering approach for gravity recovery and climate experiment (GRACE) gravity data. J Geophys Res Solid Earth 113(B4). doi:10.1029/ 2007JB005043Google Scholar
- Flechtner F, Dobslaw H (2014) AOD1B product description document for product release 05 (Rev. 4.2, May 20, 2014). Technical Note, GFZ German Research Centre for Geosciences Department 1: Geodesy and Remote SensingGoogle Scholar
- Frappart F, Ramillien G, Biancamaria S, Mognard NM, Cazenave A (2006) Evolution of high-latitude snow mass derived from the GRACE gravimetry mission (2002–2004). Geophys Res Lett 33(2). doi:10.1029/2005GL024778Google Scholar
- Jekeli C (1981) Alternative methods to smooth the Earth’s gravity field. Ohio State University, Department of Geodetic Science and Surveying, Report No. 327Google Scholar
- Landerer FW, Dickey JO, Güntner A (2010) Terrestrial water budget of the Eurasian pan-arctic from GRACE satellite measurements during 2003–2009. J Geophys Res Atmos 115(D23). doi:10.1029/2010JD014584Google Scholar
- Milly P, Betancourt J, Falkenmark M, Hirsch R, Kundzewicz Z, Lettenmaier D, Stouffer R (2008) Stationarity is dead: whither water management? Earth 4:20. doi:10.1126/science.1151915Google Scholar
- Ogawa R (2010) Transient, seasonal and inter-annual gravity changes from GRACE data: geophysical modelings. Ph.D. thesis, Department of Natural History Sciences, Graduate School of Science, Hokkaido UniversityGoogle Scholar
- Ray RD, Rowlands DD, Egbert GD (2003) Tidal models in a new era of satellite gravimetry. Space Sci Rev 108(1–2):271–282. doi:10.1023/A: 1026223308107Google Scholar
- Sasgen I, Konrad H, Ivins ER, Van den Broeke MR, Bamber JL, Martinec Z, Klemann V (2013) Antarctic ice-mass balance 2003 to 2012: regional reanalysis of GRACE satellite gravimetry measurements with improved estimate of glacial-isostatic adjustment based on GPS uplift rates. Cryosphere 7(5):1499–1512. doi:10.5194/tc-7-1499-2013CrossRefGoogle Scholar
- Schmidt R, Petrovic S, Güntner A, Barthelmes F, Wünsch J, Kusche J (2008) Periodic components of water storage changes from GRACE and global hydrology models. J Geophys Res Solid Earth 113(B8). doi:10.1029/ 2007JB005363Google Scholar
- Velicogna I, Tong J, Zhang T, Kimball JS (2012) Increasing subsurface water storage in discontinuous permafrost areas of the Lena river basin, Eurasia, detected from GRACE. Geophys Res Lett 39(9). doi:10.1029/ 2012GL051623Google Scholar
- Zhang Z-Z, Chao BF, Lu Y, Hsu H-T (2009) An effective filtering for grace time-variable gravity: fan filter. Geophys Res Lett 36(17). doi:10.1029/ 2009GL039459Google Scholar
- Zhong M, Duan J, Xu H, Peng P, Yan H, Zhu Y (2009) Trend of China land water storage redistribution at medi- and large-spatial scales in recent five years by satellite gravity observations. Chin Sci Bull 54(5):816–821. doi:10.1007/s11434-008-0556-2Google Scholar