Research ArticleENVIRONMENTAL SCIENCES

Arctic sea ice, Eurasia snow, and extreme winter haze in China

+ See all authors and affiliations

Science Advances  15 Mar 2017:
Vol. 3, no. 3, e1602751
DOI: 10.1126/sciadv.1602751
  • Fig. 1 PM pollution and ventilation conditions over East Asia.

    (A) 2013 monthly Moderate Resolution Imaging Spectroradiometer (MODIS) AOD (unitless) at 550 nm onboard Aqua satellite; (B) time series of aerosol observations, PPI, and CFI; their correlation coefficients (r values) with PPI are shown in parentheses; (C) 2013 distributions of normalized surface WSI (unitless); (D) 2013 distributions of normalized potential ATGI (unitless). In (C) and (D), black dots (crosses) denote the 99% (95%) significance level based on the bootstrapping method. The red rectangular box in (A) and the black box in (C) and (D) show the ECP region. All results are for January.

  • Fig. 2 Influence of the regional circulation on PPI.

    (A) The first MCA mode intensity (circles; unitless) of January PPI and Z850; color shading (unitless) denotes PPI values from 1981 to 2015; (B) the spatial pattern of the first Z850 MCA mode (color shading; unitless) and Z850 climatology (contour lines; in meters); (C) the spatial pattern of the first PPI MCA mode (color shading; unitless) and PPI fields (contour lines; unitless) in 2013; (D) the 2013 Z850 anomalies (color shading; in meters) and Z850 climatology (contour lines; in meters). In (D), black dots denote the 95% significance level based on the bootstrapping method; green rectangles denote the regions of subplots (B) and (C); H and L indicate the location of the Siberian High and the Aleutian Low, respectively. All results are for January.

  • Fig. 3 PPI responses to cryosphere forcing in the CESM sensitivity simulations.

    (A) The CDF of PPI over the ECP region in the reanalysis data and CESM sensitivity simulations. The red solid vertical line is at the PPI value of 0; (B) the spatial distribution of ensemble averaged PPI fields of the extreme members (Embedded Image) in SENS1; (C) same as (B) but in SENS2; (D) same as (B) but in SENS3. In (B) to (D), black dots denote the 99% significance level based on the bootstrapping method.

  • Fig. 4 Time series of cryospheric forcing factors, PPI, and the correlations of PPI with cryospheric forcing based on the observations, reanalysis data, and CMIP5 projections.

    (A) Comparisons of SIC observations and ensemble averaged CESM1(CAM5) (three ensemble members) and CCSM4 (six ensemble members) CMIP5 simulations (unitless; color shading denotes ±1 SD); (B) same as (A) but for SCE; (C) same as (A) but for CFI; (D) same as (A) but for PPI; (E) correlation coefficients of PPI with SCE, SIC, and CFI based on the 35-year observations and reanalysis data [symbols; unitless; the absolute value of the PPI-SIC correlation coefficient (magenta square) is shown for comparison purpose] and 35-year moving correlation coefficients between simulated PPI and CFI of CMIP5 ensembles (lines; unitless; the data are plotted at the last year of each 35-year period).

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/3/3/e1602751/DC1

    fig. S1. PCA decomposition and reconstruction of PPI.

    fig. S2. Cryosphere forcing specifications used in the CESM numerical experiments.

    fig. S3. Sensitivity response of the first coupled PPI and Z850 modes to cryospheric forcing.

    fig. S4. Comparison of surface air temperature in December between reanalysis data and numerical experiments.

    fig. S5. Time series of monthly WSI, ATGI, and PPI over the ECP region for January.

    table S1. A list of acronyms used in this study.

    table S2. Correlations among ventilation indices, PM observations, and cryospheric forcing factors.

    table S3. Climate and synoptic weather indices in the PCA.

    table S4. Correlation statistics among PCs and climate indices.

    table S5. PCR coefficients of the detrended PPI onto PC.

    table S6. The CMIP5 ensemble simulations used in this study.

    Reference (44)

  • Supplementary Materials

    This PDF file includes:

    • fig. S1. PCA decomposition and reconstruction of PPI.
    • fig. S2. Cryosphere forcing specifications used in the CESM numerical experiments.
    • fig. S3. Sensitivity response of the first coupled PPI and Z850 modes to cryospheric forcing.
    • fig. S4. Comparison of surface air temperature in December between reanalysis data and numerical experiments.
    • fig. S5. Time series of monthly WSI, ATGI, and PPI over the ECP region for January.
    • table S1. A list of acronyms used in this study.
    • table S2. Correlations among ventilation indices, PM observations, and cryospheric forcing factors.
    • table S3. Climate and synoptic weather indices in the PCA.
    • table S4. Correlation statistics among PCs and climate indices.
    • table S5. PCR coefficients of the detrended PPI onto PC.
    • table S6. The CMIP5 ensemble simulations used in this study.
    • Reference (44)

    Download PDF

    Files in this Data Supplement:

Related Content

More Like This