Papers of the Month 【201709】

温馨提示：框内图片如加载缓慢，可以点击图片区域直接浏览。

1A Direct Estimate of the Seasonal Cycle of Evapotranspiration over the Amazon Basin

By: Abigail L. S. Swann, Charles D. Koven

Journal: Journal of Hydrometeorology

Abstract:

Fig. 1. Precipitation and basin used in this analysis. Maps showing (a) the annual mean precipitation (mm yr^-1) and (b) the length of the dry season (months) where blue colors indicate wetter conditions and red and brown colors indicated drier conditions. An outline of the basin used in this analysis—the Amazon basin upstream of the gauging station at Óbidos—is shown in the black outline. Red markers indicate the location of four flux towers near the region from Wu et al. (2016).

Fig. 2. Comparison between terms derived from S using monthly mean values and those derived using storage interpolated to a daily scale using a spline. (a) Parameter S as directly reported monthly means (blue) vs interpolated to daily values using a spline (red); (b) the derivative of S (dS/dt) taken using the derivative of the daily interpolated S (cyan) and via a centered finite difference of the monthly values (darker blue); and (c) the estimated ET from the water budget approach using rainfall fromTRMM, with dS/dt de-rived using daily interpolated values of S (darker green) vs dS/dt derived via a centered finite difference of monthly mean values of S (lighter green).

Fig. 3. Full time series and seasonal climatology of the Amazon basin–averaged water budget terms (mm yr^-1). The (a) time series and (b) seasonal climatology of each component of the water budget and the estimated ET calculated from Eq. (1) (ET_GRACE), where precipitation and its range is shown in gray-blue, the change in storage over time (dS/dt) is shown in bright blue, runoff is shown in brown, and ET_GRACE and its range is shown in gray.

Fig. 4. Comparison of the climatological seasonal cycle of estimated ET from several sources. (a) The range of estimates of ETGRACE using four precipitation datasets in different colors (TRMM, GPCP, GPCC, and CRU) and two finite difference approaches, with ET derived from a finite backward difference of storage interpolated to a daily scale (solid lines) and derived using a finite centered difference using monthly mean values of storage (dashed lines). Gray shading indicates the total range encompassed by all eight estimates. The climatological seasonal cycle of ET_GRACE (gray line, with range in gray shading) is compared with (b) other ET products and (c) models. (d) Comparison of ET_GRACE (gray) with flux-tower-based measurements and products including theFLUXNET-MTE product (mauve) and four flux towers near the region from Wu et al. (2016) (locations shown in Fig. 1). (e) The normalized seasonal cycle (in units of stddev) for each of the climatologies shown in (b)–(d).

Fig. 5. Time series of all available data between 2002 and 2016 for ET estimated using four different categories of methods. (a) ET estimated using a water budget approach (ET_GRACE; solid line), estimated range of uncertainty (gray shading), and Monte Carlo–estimated regression (slope of -1.46 mm yr^-1,CI from -2.4 to -0.51 mm yr^-1). (b) ET estimated using several different energy budget type approaches (see legend). (c)ET estimated from site-level data from eddy covariance towers upscaled with satellite data. (d) ET estimated from land surface models forced with the same meteorology from the TRENDY model project.

Fig. 6. ET estimates for all years of available data for ET_GRACE (gray lines and shading) compared with upscaled site-level data (a)FLUXNET-MTE, and energy budget products (b) MOD16, (c) ET-M, (d) BESS, (e) P-LSH, and (f) JF-ET.

Fig. 7. The seasonal cycle of ET_GRACE (gray line and shading) compared with the seasonal cycle in downwelling shortwave radiation (SW down; orange) and SIF (green). Estimated potential ET based on SW down (orange) is shown by the rightmost y axis.

2A Global Gridded Dataset of GRACE Drought Severity Index for 2002-2014: Comparison with PDSI and SPEI and a Case Study of the Australia 4 Millennium Drought

By: Meng Zhao, Geruo A, Isabella Velicogna, John S. Kimball

Journal: Journal of Hydrometeorology

Abstract:

Fig. 1 Global patterns 29 40867 29 12198 0 0 8090 0 0:00:05 0:00:01 0:00:04 8104 of the GRACE-DSI, PDSI-Z and SPEI-Z drought metrics at selected timescales (1, 3, 6, 9, 12, 15, 24, and 36 months) for July 2010.

Fig. 2 Time series of the GRACE-DSI (red), PDSI-Z (black), and SPEI-Z (blue) at selected timescales at four locations annotated with 350km radius footprints in the top panel. The geographic coordinates are (46E, 54N), (72W, 8S), (38E, 0S), and (116E, 44N) for locations 1 to 4, respectively. Note that the GRACE-DSI are the same for all plots in the same location. Error bars on GRACE-DSI represent the GRACE-DSI uncertainty due to GRACE measurement and leakage errors. Pearson correlation coefficient of each comparison is shown on top of each plot. Correlation coefficient larger than 0.17 is significant at 95% confidence level.

Fig. 3 (a) is the correlation between monthly GRACE-DSI and PDSI-Z. (b) – (i) are correlations between monthly GRACE-DSI and SPEI-Z at time-scales of 1, 3, 6, 12, 18, 27, 36, and 48 months, respectively. (j) is the maximum correlation between monthly GRACE-DSI and SPEI-Z at various time scales. (k) is the time scale of SPEI-Z in which the maximum correlation in (j) is recorded. Insignificant correlation coefficients (p>0.05) are masked out in panels (a-j).

Fig. 4 Time series of GRACE-DSI (red), satellite-retrieved SM-Z (yellow), and NDVI-Z (green) for two locations in mainland Australia annotated with 350km footprints in the land cover map (a). (b) is for location 1 (27S, 121E) in west Australia and (c) is for location 2 (28S, 148E) in east Australia. Time series are smoothed using a quadratic polynomial filter with a 13-month window (Savitzky and Golay 1964). Uncertainties of these satellite records are shaded in corresponding colors. The errors of SM-Z and NDVI-Z are estimated conservatively in a similar manner as the GRACE-DSI considering both measurement error and leakage error.

Fig. 5 Global distribution of GRACE-DSI uncertainty in drought category excluding the Antarctic, Greenland, and barren grounds.

Fig.6 (a) is the time scale by which NDVI-Z achieves maximum correlation coefficient with SPEI-Z. (b) is the Australia sub-region in Fig. 3k, i.e., the time scale by which GRACE-DSI achieves maximum correlation coefficient with SPEI-Z. Note that a large area of (b) saturates at time scales over 20-months. Corresponding maximum correlation coefficients are significant at 99% confidence level for both plots.

Fig.7 Drought category overestimation (positive value) and underestimation (negative value) using the 2002-2014 reference period rather than the1982-2014 reference period for the PDSI-Z drought index.

3Evaluating Renewable Groundwater Stress with GRACE Data in Greece

By: Alexandra Gemitz, Venkat Lakshmi

Journal: Groundwater

Abstract:

Fig. 1. Comparison of GRACE-derived TWS anomalies to SWAT modeled TWS anomalies from January 2005 to July 2015 (a) in Thrace region and (b) in Thessaly region.

Fig. 2. Water storage components in Thrace region. (a) Total GRACE-derived water storage anomalies, (b) SWS anomalies from SWAT model, (c) SOWS anomalies from SWAT model, (d) subsurface storage anomalies as the difference between total storage anomalies and the sum of SWS and SOWS. p-values correspond to the statistical significance of trends.

Fig. 3. Water storage components in Thessaly region. (a) Total GRACE-derived water storage anomalies, (b) SWS anomalies from SWAT model, (c) SOWS anomalies from SWAT model, (d) subsurface storage anomalies as the difference between total storage anomalies and the sum of SWS and SOWS. p-values correspond to the statistical significance of trends.

Fig. 4. GRACE-derived GWS in the aquifers of Greece.

Fig. 5. Mean annual groundwater recharge rates in the aquifers of Greece.

Fig. 6. RGS in the aquifers of Greece.

4Glacier Mass Balance and Its Potential Impacts in the Altai Mountains over the Period 1990–2011

By: Yong Zhang, Hiroyuki Enomoto, Tetsuo Ohata, et al

Journal: Journal of Hydrology

Abstract:

Fig. 1 Location of the Altai Mountains and glacier distribution. Abbreviations refer to: AKT, Aktru River Basin (including Leviy Aktru (LA), Maliy Aktru (MA), and Praviy Aktru (PA) and Vodopadniy (VP) glaciers); PTN, Potanin Glacier; TGV, Tsambagarav Glacier. No. 1-10 show the weather stations, which are Habahe, Jimunai, Aletai, Fuhai, Fuwen, Qinghe, Ust-Coksa, Kara-Tureck, Kosh-Agach and Mugur-Aksy, respectively.

Fig. 2 (a) Area-altitude distribution of the Altai glaciers, (b) glacier distribution for different area size classes and (c) mass balance and sensitivities to temperature and precipitation changes for different area size classes.

Fig. 3 Scatter plots of daily (a and b), monthly (c and d), and annual (e and f) temperature and precipitation between the WRF simulation and station observation.

Fig. 4 Seasonal temporal trend (black) and seasonal variation (grey) of temperature (a) and precipitation (b) during 1990–2011 estimated from ten weather stations (OBS) and the WRF simulations (WRF).

Fig. 5 (a) Scatter diagram of observed and modelled annual mass balance for the corresponding observation period, (b) time series of observed and modelled annual mass balance in the Aktru River Basin over the period 1990–2011, (c) geodetic mass change rates and modelled results on the fourteen glaciers in the Chinese Altai Mountains, and (d) satellite-based observed and modelled glacier area change in the Aktru River Basin over the period 1999–2008 and the Chinese Altai Mountains over the period 2000–2008. Light gray shading in (b) denotes the standard deviation of the estimated mass balance, and error bar in (c and d) indicates standard error. LA, MA, PA and VP denote the four glaciers in the Aktru River Basin and W01–W14 denote the fourteen glaciers in the Chinese Altai Mountains and their information can be found in Wei et al. (2015).

Fig. 6 (a) Time series of the modelled annual and cumulative mass balance of the entire glaciers in the Altai Mountains during 1990–2011, and (b) observed annual mass balance of different glaciers during 1977–2012. Light gray shading in (a) denotes the standard deviation of the estimated mass balance.

Fig. 7 Spatial variability of modelled surface mass balance of the glaciers in the Altai Mountains averaged during 1990–2011. The bottom right shows glacier mass balance in the southeastern part of the Mongolian Altai Mountains.

Fig. 8 Spatial patterns of mean annual air temperature (a), trend in temperature change (b), annual precipitation (c), and trend in precipitation change (d) over the period 1990–2011 on each glacier in the Altai Mountains calculated from WRF data. The bottom right in each figures shows the spatial patterns on each glacier in the southeastern part of the Mongolian Altai Mountains. 

Fig. 9 Contribution to sea-level rise from the entire glaciers of the Altai Mountains during different periods.

Fig. 10 Spatial variability of mass balance sensitivities of the entire glaciers to changes in (a) temperature and (b) precipitation in the Altai Mountains. The bottom right in (a) and (b) shows temperature and precipitation sensitivities in the southeastern part of the Mongolian Altai Mountains.

Fig. 11 Influences of changing in (a) the DDFs for snow and ice and (b) temperature lapse rates and precipitation gradients on modelling mass balance in the Aktru River Basin. Here, the DDFs for snow and ice changes by ±10% (DDF±10%) and ±30% (DDF±30%), respectively, and temperature lapse rates and precipitation gradients changes by ±10% (TP±10%), ±30% (TP±30%) and ±50% (TP±50%), in which other parameters remain the same as before (Modelled).

5Groundwater-level Trends in the U.S. Glacial Aquifer System, 1964-2013

By: G.A. Hodgkins, R.W. Dudley, M.G. Nielsen, et al

Journal: Journal of Hydrology

Abstract:

Fig. 1. Geographic distribution of trends in mean annual groundwater levels, 1984-2013; (a) all qualifying wells (including wells with both low human-influence potential and high human- influence potential) and assuming data series independence; (b) wells with low influence potential and assuming data series independence; (c) wells with low influence potential and assuming data series short-term persistence; (d) wells with low influence potential and assuming data series long-term persistence.

Fig. 2. Geographic distribution of trends in mean annual groundwater levels, 1964-2013; (a) all qualifying wells (including wells with both low human-influence potential and high influence potential) and assuming data series independence; (b) wells with low influence potential and assuming data series short-term persistence.

Fig. 3. Geographic distribution of trends in annual precipitation, 1984-2013; (a) assuming data-series independence; (b) assuming data-series short-term persistence; (c) assuming data-series long-term persistence.

Fig. 4. Geographic distribution of trends in annual precipitation, 1964-2013; (a) assuming data series independence; (b) assuming data series short term persistence; (c) assuming data series long term persistence.

Fig. 5. Distribution of mean groundwater-level trends, 1994-2013; (a) East Region, by wells with low human-influence potential, high influence potential (from urban and/or agricultural influences), and high urban-influence potential but low agricultural-influence potential; (b) Central Region, by wells with low human-influence potential, high influence potential, high urban-influence potential but low agricultural-influence potential, and high agricultural-influence potential but low urban-influence potential.

Fig 6. Plot of mean annual precipitation by region, 1944-2013. Lines are locally weighted regression (LOESS) with a weighting factor of 0.33.

Fig. 7. Mean annual groundwater levels from a U.S. Geological Survey well in north-central Massachusetts, 1944-2013. Line is locally weighted regression (LOESS) with a weighting factor of 0.33.

6Large-scale Vegetation Responses to Terrestrial Moisture Storage Changes

By: Robert L. Andrew, Huade Guan, Okke Batelaan

Journal: Hydrology and Earth System Sciences

Abstract:

Fig. 1. (a) The spatial distribution of various land use types across Australia and (b) the area covered by each land use type.

Fig. 2. An example of a wavelet decomposition from a cell in central South Australia (29^◦ S, 136^◦ E). Note the visible trends in the approximations (A1–A4), which are normalised in the details (D1–D4).Fig. 3. The structure of a wavelet decomposition; decomposition levels used in this study are highlighted in red.Fig. 4. (a) The r values of the relationship between the raw TWS^∗ and NDVI^∗. (b) The r values of the relationship between the de-composed TWS^∗ and NDVI^∗. (c) A scatter plot of the r values of both relationships shows a clear improvement in the results when the decomposed data are used.Fig. 5. An example of the time series from a single cell. The new estimate uses the coefficients from A4₀, A4₆, and D4 as selected by the stepwise regression. Pearson’s coefficient (r) between the decomposed 

GRACE estimate and NDVI^∗ 

is 0.872, compared with 0.665 when using raw GRACE TWS^∗.Fig. 6. Patterns of the coefficients for each decomposition level. For D1 and D2, no lags are used; for these, red represents a positive coefficient and blue represents a negative coefficient. For D3_L, D4_L, and A4_L (which include lags), red represents cells where all coefficients are positive and blue represents cells where at least one lag had a negative coefficient.Fig. 7. The variable with the highest relative weight in the regression for each cell across Australia. A4 is the most dominant; however D2 is prominent in distinct areas throughout central Australia. D1, D3_L, and D4_L all occur but with little spatial coherence.Fig. 8. The relative weight of each decomposed TWS^∗ for each land use type. Forests are A4_L dominated; shrublands, savannas and grasslands are very similar with relative equal weights of D1, D2, and A4_L and agricultural land is dominated by D2 and A4_L.7Long-Term Water Storage Changes of Lake Volta from GRACE and Satellite Altimetry and Connections with Regional Climate

By: Shengnan Ni, Jianli Chen, Clark R. Wilson, Xiaogong Hu

Journal: Remote Sensing

Abstract:

Fig. 1. The map of Volta River basin in West Africa. Original map adapted from http://www.zef.de/publ_maps.html.

Fig. 2. Total water storage changes from the Gravity Recovery and Climate Experiment (GRACE) over the Volta River basin as outlined in Fig. 1.

Fig. 3. Principal component analysis (PCA)-derived spatial and temporal patterns of terrestrial water storage (TWS) variability (with annual and semiannual signals removed) over the Volta River basin. (a,b) are spatial patterns of the first two modes derived from PCA; (c,d) are corresponding temporal patterns. The percentages of the total variance explained by the first two principal components are 30.9% and 20.7%, respectively.

Fig. 4. GRACE water storage changes and satellite altimetry water level changes for Lake Volta. Both time series in (a,b) have an increasing (2007–2010) and declining (2011–2015) rate; (c) is the comparison between GRACE and satellite altimetry at long-term time scale. We have removed the annual and semi-annual signals using least squares fitting. Please notice the different y-axis scales used in (c).

Fig. 5. Global Precipitation Climatology Centre (GPCC) monthly precipitation over the Volta River basin with the climatologic average removed. The climatological precipitation is calculated by averaging the monthly precipitation of all the same months over a certain period (e.g., the 20-year period from January 1996 to December 2015). The red line is the nonseasonal precipitation anomaly smoothed with a Butterworth low-pass (below 0.5 cpy) filter.

Fig. 6. GRACE TWS long-term change rates and GPCC mean precipitation anomalies over the Volta River basin during the periods of 2007–2010 and 2011–2015. (a) is TWS long-term change rates from 2007 to 2010 after P4M6 decorrelation filtering and 300 km Gaussian smoothing; (b) is mean precipitation anomalies from 2007 to 2010 without any smoothing filter; (c) is TWS long-term change rates from 2011 to 2015 after P4M6 decorrelation filtering and 300 km Gaussian smoothing; (d) is mean precipitation anomalies from 2011 to 2015 without any smoothing filter. Mean precipitation anomalies are the average values of precipitation with a certain period (e.g., 20 years) climatology removed.

Fig. 7. Mass rates (January 2007–December 2010) in cm/year of equivalent water height. (a) Apparent long-term TWS change rates from GRACE after P4M6 decorrelation filtering and 300 km Gaussian smoothing; (b) Restored “true” long-term TWS change rates from constrained forward modeling after 300 iterations; (c) Predicted TWS change rates from model rates of (b); (d) Difference between observed and modeled apparent mass rates (i.e., (a–c)). Please notice the different color scale used in the four panels.

Fig. 8. Mass rates (January 2011–December 2015) in cm/year of equivalent water height. (a) Apparent long-term TWS change rates from GRACE after P4M6 decorrelation filtering and 300 km Gaussian smoothing; (b) Restored “true” long-term TWS change rates from constrained forward modeling after 200 iterations; (c) Predicted TWS change rates from model rates of (b); (d) Difference between observed and modeled apparent mass rates (i.e., (a–c)). Please notice the different color scale used in the four panels.

Fig. 9. Residuals between observed and apparent mass rates in constrained forward modeling. The residual is computed as the root mean square (RMS) value of difference between observed and modeled data at each grid point over the entire rectangle region shown in Fig. 7.

Fig. 10. Residuals between observed and apparent mass rates in Fig. 8.

Fig. 11. GRACE water storage changes (equivalent water volume) with leakage correction and satellite altimetry water volume changes for Lake Volta. The red curve can be obtained by multiplying the red curve in Figure 4c with both scale factor (~41.3) and the area of lake mask. The blue curve is the product of altimetry water level change (blue curve in Figure 4c) with estimated lake area. Please notice that the y-axis scale in this figure is the same for GRACE and satellite altimetry data.

8Regional Characteristics and Influencing Factors of Seasonal Vertical Crustal Motions in Yunnan, China

By: Wei Zhan, Fei Li, Weifeng Hao, Jianguo Yan

Journal: Geophysical Journal International

Abstract:

Fig. 1. Schematic distribution of continuous GPS stations (triangles) and rainfall stations (circles) in Yunnan. Red solid line represents Jinshajiang–Red River Fault (Deng et al. 2003). Fig. 1(b) shows the location of study area in the Tibetan Plateau, the background shows topography.

Fig. 2. Annual vertical motion in Yunnan during 2010–2015 with the reference time as 2010.0. The arrows denote the amplitudes, the azimuthal angles counted counterclockwise from the east represent the initial phase of the sinusoidal function. The arrows pointing to the east, south, west and north indicate that the maximum occurs at March, June, September and December, respectively. The annual motions of GPS stations are shown as black arrows, and annual motion of regional common-mode component (CMC) is shown as blue arrow. The error ellipses represent 67 per cent confidence.

Fig. 3. The WRMS reduction ratios for all sites. The colour bar is shown at bottom.

Fig. 4. Correlation and phase relationships between monthly average of detrended GPS vertical time-series and monthly sum precipitation. Circles show that the station locations. Circle colour (shown at bottom) gives the correlations, there is a direct inverse correlation between rainfall and displacements. The number gives the phase difference, usually by 2 months.

Fig. 5. Monthly average of common-mode component (MACMC, red circle, a) and average of monthly sump precipitation (AMSP, b).

Fig. 6. GPS-observed vertical velocities in Yunnan (2010–2015). The error bars represent 67 per cent confidence.

9Statistically Optimal Estimation of Degree-1 and C20 Coefficients Based on GRACE Data and an Ocean Bottom Pressure Model

By: Yu Sun, Pavel Ditmar, Riccardo Riva.

Journal: Geophysical Journal International.

Abstract:

Fig. 1. Oceanic mass anomalies updates at different iterations. In panels (a)–(d), we show the updates for iteration 1 (RMS: 2 mm; maximum value: 6 mm), iteration 2 (RMS: 0.4 mm; maximum value: 0.8 mm), iteration 3 (RMS:0.08 mm, maximum value: 0.2 mm) and iteration 4 (RMS: 0.01 mm, maximum value: 0.03 mm), respectively.

Fig. 2.Uncertainty of oceanic mass anomalies. (a) The RMS error of OBP predictions. (b) The RMS error of fingerprints. (c) Total RMS error obtained with eq. (9).

Fig. 3.Actual error, actual RMS error (ARE) and approximated actual RMS error (approx. ARE) for one realization of synthetic C₁₀ coefficients. The actual errors are obtained as the differences between the resulting C₁₀ time-series and the synthetic truth. ARE is then obtained as the RMS difference, which is shown as a grey band (the upper and lower bound of the grey band is then ± ARE).

Fig. 4. Combined ARE, combined FRE and combined approximated ARE as functions of the scaling factor. The grey band along the combined approximated ARE curve shows its STD from 10 GSM realizations. Note that the parameter setup is as follows: truncation degree: 50, buffer zone width: 200 km.

Fig. 5. The RMS errors (average over 10 simulated GSM solutions) for resulting degree-1 and C₂₀ coefficient time-series (in mm of equivalent water height). Results for C₁₀, C₁₁, S₁₁ and C₂₀ are presented in panels (a)–(d), respectively. The thick grey lines show the results of the GRACE-OBP-Swenson approach. The dashed grey lines indicate solutions based on the GRACE-OBP-Improved approach considering SAL effects and using a 200 km buffer zone.

Fig. 6.The same as Fig. 5, but the unified quality indicator is shown instead of RMS errors per coefficient.

Fig. 7.The mean annual amplitudes of the GSM degree-1 and C₂₀ time-series (mm EWH) estimated using different implementation parameters, based on 10 sets of simulated GSM solutions. The standard deviations of amplitude estimates (based on 10 sets of GSM solutions) are indicated by light coloured bands. The true amplitudes are marked in all panels as black horizontal lines. Results for C₁₀, C₁₁, S₁₁ and C₂₀ are shown in panels (a)–(d), respectively.

Fig. 8. The same as Fig. 7, but for the annual phases.

Fig. 9. The same as Fig. 6, but the quality indicator for annual variations is shown instead of quality indicator for RMS error.

Fig. 10.Combined FRE and combined approximated ARE as functions of the scaling factor. Note that the parameter setup is as follows: truncation degree: 50, buffer zone width: 200 km.

Fig. 11. Final solutions for degree-1 and C₂₀ time-series (a), their formal error estimates (b) and correlation coefficients (c) based on the combination approach. In panel (a), linear trends are removed. Results are offset for clarity. The coloured bands show the 2-σ uncertainties. The black dashed line shown in panel (b) (denoted as ‘Swenson et al.’) is taken from the official product based on (Swenson et al. 2008) (see the text).

Fig. 12. Mass variations in the validation area at East Antarctica. In panel (a), we show the signal RMS in terms of equivalent water heights. The study area is indicated with a red polygon (45^◦E/120^◦E/76^◦S/84^◦S). Panel (b) shows RMS values of mass anomaly time-series as a function of truncation degree. Panels (c) and (d) show the mean mass anomaly per calendar month. The shadowed colour bands indicates the spread of the monthly mass anomalies. Note that the calendar month 0 represents December of the previous year. In panel (c), we show the results based on the GRACE solutions after replacing the C₂₀ coefficients with those from independent approaches. In panel (d), we show results when the GRACE solutions are further complemented with degree-1 coefficients based on different approaches.

Fig. 13.Same as Fig. 12, but showing mass transport in the validation area at the Sahara Desert (6^◦W/30^◦E/15^◦N/30^◦N).

卫星重力与水文学

长按二维码关注我们吧！

信息发布|学术交流|资讯共享

Email: [e-mail]