Research ArticleENVIRONMENTAL STUDIES

Agreement between reconstructed and modeled boreal precipitation of the Last Interglacial

See allHide authors and affiliations

Science Advances  20 Nov 2019:
Vol. 5, no. 11, eaax7047
DOI: 10.1126/sciadv.aax7047
  • Fig. 1 Zonal mean anomalies between LIG and PI climates for each climate model and the ensemble mean.

    Anomalies of near-surface air temperature over land and ocean (top) and precipitation over land only (bottom) are shown for the annual mean and each season (defined after the angular calendar). Annual and seasonal insolation anomalies [between 127 ka and 1850 CE; (6)] are plotted as dashed lines.

  • Fig. 2 Annual precipitation anomaly between LIG and PI from models (ensemble average in contoured colors) and proxies (filled markers)

    Blue colors indicate higher precipitation during the LIG from models or proxies and conversely for red colors. Proxy anomalies are on a semiquantitative scale (see Methods): dark blue (much wetter LIG), light blue (wetter), white (no noticeable anomaly), light red (drier), and dark red (much drier). Different markers represent different types of proxy records. Hatched areas indicate regions where one (single hatching) or at least two (double hatching) of seven models disagree with the sign of the ensemble average anomaly.

  • Fig. 3 Percentage anomalies between LIG and PI.

    Left: Anomalies in area-weighted precipitation. Right: Anomalies in monsoon domain area and monsoon precipitation for each monsoon and each hemisphere. Monsoon domains were calculated after Wang and Ding (43) and are reported in fig. S5. Monsoon precipitation is calculated within the monsoon domains. Only grid cells over land are considered.

  • Fig. 4 Precipitation anomaly between LIG and PI, comparison of models and proxies.

    The hit rate is the percentage of matches between the sign of the anomaly in the model ensemble average and in the proxy records in the database (see Methods). n is the number of comparisons per region.

  • Fig. 5 Model ensemble average of seasonal precipitation anomalies between LIG and PI.

    Seasons are defined after the angular calendar. Hatched areas are as in Fig. 2.

  • Fig. 6 Model ensemble average of seasonal precipitation anomalies between LIG and PI.

    Top four panels: Precipitation extremes (RX5day index, i.e., maximum precipitation during five consecutive days, in millimeters). Hatched areas are as in Fig. 2. Bottom four panels: Synthesis of climatological average precipitation and precipitation extremes. Green indicates more average precipitation during the LIG. Blue indicates more precipitation extremes during the LIG. Purple indicates both more average precipitation and more precipitation extremes during the LIG.

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/5/11/eaax7047/DC1

    Supplementary Text S1. Temperature anomalies from models.

    Supplementary Text S2. Comparison of seasonal precipitation between proxies and models.

    Supplementary Text S3. Limits of proxy dataset and of proxy-model comparisons.

    Fig. S1. Anomaly in insolation between the LIG (127 ka) and the PI (year 1850 CE).

    Fig. S2. Comparison of annual anomalies between LIG and PI.

    Fig. S3. Ensemble average anomaly between LIG and PI for variables diagnostic of atmospheric circulation.

    Fig. S4. Difference between seasonal ensemble anomalies (between LIG and PI) after the angular and the classic calendar.

    Fig. S5. Monsoon Precipitation Index, in contoured colors, and monsoon domains, in contoured black lines.

    Fig. S6. Approximate position of the ensemble mean seasonal intertropical convergence zone (over the oceans) and tropical rain belt (over the continents).

    Fig. S7. Model ensemble near-surface air temperature anomalies between LIG and PI.

    Fig. S8. Correlations between LIG-PI anomaly in average precipitation and in extreme precipitation, per season, for the ensemble average.

    Fig. S9. Selected sets of correlations between anomalies in temperature and North African monsoon for each model in the ensemble.

    Fig. S10. Precipitation from the ensemble mean of the PI and LIG simulations.

    Fig. S11. Bias of the PI ensemble annual average precipitation with respect to the closest comparable reanalysis products.

    Table S1. Details of the models in the ensemble, with main information regarding the modules relevant to this study.

    External database S1. Proxy database.

  • Supplementary Materials

    The PDFset includes:

    • Supplementary Text S1. Temperature anomalies from models.
    • Supplementary Text S2. Comparison of seasonal precipitation between proxies and models.
    • Supplementary Text S3. Limits of proxy dataset and of proxy-model comparisons.
    • Fig. S1. Anomaly in insolation between the LIG (127 ka) and the PI (year 1850 CE).
    • Fig. S2. Comparison of annual anomalies between LIG and PI.
    • Fig. S3. Ensemble average anomaly between LIG and PI for variables diagnostic of atmospheric circulation.
    • Fig. S4. Difference between seasonal ensemble anomalies (between LIG and PI) after the angular and the classic calendar.
    • Fig. S5. Monsoon Precipitation Index, in contoured colors, and monsoon domains, in contoured black lines.
    • Fig. S6. Approximate position of the ensemble mean seasonal intertropical convergence zone (over the oceans) and tropical rain belt (over the continents).
    • Fig. S7. Model ensemble near-surface air temperature anomalies between LIG and PI.
    • Fig. S8. Correlations between LIG-PI anomaly in average precipitation and in extreme precipitation, per season, for the ensemble average.
    • Fig. S9. Selected sets of correlations between anomalies in temperature and North African monsoon for each model in the ensemble.
    • Fig. S10. Precipitation from the ensemble mean of the PI and LIG simulations.
    • Fig. S11. Bias of the PI ensemble annual average precipitation with respect to the closest comparable reanalysis products.
    • Table S1. Details of the models in the ensemble, with main information regarding the modules relevant to this study.
    • Legend for external database S1

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article