Figure 1. The rainfall estimates in the upper panel are a hybrid of two different datasets. Over the oceans they are based on the MSU carried aboard the Television Infrared Operational Satellites (TIROS) for the period 1979-1991. The method for inferring rainfall from MSU measurements is described in Spencer (1993). Over land, they are based on a compilation of historical raingauge data by Legates and Willmott (1990). The GPI is described in Arkin and Meisner (1987), and based on 1986-1993. MSU data provided by the NASA Marshall Space Flight Center.
Figure 2. Surface winds are based on an analysis of the COADS (Woodruff et al. 1987, 1993) performed by Sadler et al. (1987). They represent averages for the period 1900-1979. Rainfall estimates are as in the upper panel of the previous figure. The outline for the stratus cloud decks corresponds to the 0.3 contour in an albedo map derived from 4 years of visible satellite imagery. Only surface winds in excess of 3 m/s are plotted, and annual precipitation totals in excess of 2 m are shaded.
Figure 3. Surface winds and sea-surface temperature climatologies are from Sadler et al. (1987) as in the previous figure. The shading varies from blue to red for cold to warm sea surface temperatures, with gray indicating temperatures near 27°C. The wind plotting convention is as in the previous figure.
Figure 4. As in Fig. 3, but arrows denote surface currents based on the Richardson (1989) climatology of the historical record of ship drift measurements.
Figure 5. Courtesy of G. Feldman, NASA.
Figure 6. Coarse and fine resolution elevation data sets derived from the 5-minute latitude-longitude resolution TerrainBase Worldwide Digital Terrain Data Set, which is produced at the National Geophysical Data Center (NGDC) and is available online at both NGDC and NCAR.
Figure 7. Snow depth climatology from the US Air Force Environmental Technical Applications Center (Foster and Davy 1988). Model soil moisture (surface soil wetness) for 1987-88 from the ECMWF Level III-A global atmospheric data archive. FPAR observations for 1987-88 (Sellers et al. 1994, 1996). All three data sets are taken from the International Satellite Land Surface Climatology Project (ISLSCP) Initiative I CD-ROM (Sellers et al. 1996), and they are also available online through the NASA Goddard Space Flight Center Distributed Active Archive Center.
Figure 8. Circulation data taken from the NCEP/NCAR reanalysis archive. Precipitation estimates are based on merged satellite and station observations. North American panel courtesy of W. Higgins and M. Halpert; South American panel courtesy of V. Kousky and M. Halpert.
Figure 10. Figure provided by Wayne Higgins of NOAA Climate Prediction Center and Arthur Douglas of Creighton University.
Figure 11. As in Fig. 3.
Figure 12. After Mitchell and Wallace (1992), based on observations from the COADS for 1946-1985.
Figure 13. Based on data from the TOGA-TAO array (McPhaden and McCarty, 1992). The climatology for 140°W is based on observations for July 1983 to December 1991 and for 110°W from March 1980-June 1982 and July 1983-December 1991. The period of large anomalies during the 1982-1983 ENSO event, July 1982-June 1983, were not included in the climatology to prevent large biases due to this single event. Red shading for temperatures greater than 24°C, with increases in color intensity for each 1°C increase in temperature. Blue contours for temperatures less than and equal to 22°C; contour interval 2°C.
Figure 14. Derived from high-resolution infrared images from geostationary satellites, 8 per day. Courtesy of K. Howard, NOAA NSSL.
Figure 15. Surface wind changes based on the COADS; sea-level pressure based on the ECMWF operational analyses; and rainfall as in the upper panel of Fig. 1. The surface wind and sea-level pressure climatologies are based on data for 1946-1979 and 1980-1989, respectively. Only vector wind changes with magnitudes in excess of 1 m/s are plotted. Pressure changes in increments of 0.6 mb are contoured, and precipitation changes in excess of 5 cm in magnitude are shaded. The sea-level pressure analysis over land is consistent with an analysis of historical airport pressure records (not shown).
Figure 16. Water content and FPAR as in Fig. 7.
Figure 17. Snow data as in Fig. 7.
Figure 18. Rainfall as in upper panel of Fig. 1.
Figure 19. Standard deviation of monthly data for 1982-1990, provided by Paul Dirmeyer, University of Maryland. The data is from the ISLSCP Initiative II, and is available online from the NASA DAAC. The LAI is discussed in Sellers et al. (1994, 1996).
No notes for Figure 20.
Figure 21. Regression of sea surface temperature and MSU precipitation on an index of eastern equatorial Pacific (6°N-6°S, 180-80°W) sea surface temperature. The analysis is based on data for 1982-1991 and 1979-1991 for the sea surface temperature and precipitation fields, respectively. The sea surface temperature analyses are from NOAA NCEP. Temperature contour interval 0.2°C per 1 standard deviation of the index, with negative (zero) contours dashed (thickened). Shading for precipitation anomalies greater than 1 cm per 1 standard deviation of the index. Precipitation anomalies range from -4 to +10 cm per 1 standard deviation of the index.
Figure 22. Precipitation estimates from the MSU.
Figure 23. Regression of precipitation over the oceans and mean-tropospheric temperature on an index of eastern equatorial Pacific (6°N-6°S, 180-80°W) sea surface temperature. The analysis is based on data for 1979-1991. Both the precipitation and the mean tropospheric temperature fields are taken from MSU measurements. The method for inferring precipitation from MSU measurements is described in Spencer (1993) and the mean-tropospheric temperature estimates are described by Spencer and Christy (1990). Temperature contour interval of 0.1°C per 1 standard deviation of the index, with the positive and zero (negative) isotherms are drawn in gold (green). Precipitation plotting convention as in Fig. 21.
No note for Figure 24.
Figure 25. After Thompson et al. (1979).
Figure 26. The QuikSCAT wind vectors are the simple average over one degree bins and one month. The divergence is calculated from the averaged wind vectors and then zonally smoothed along each latitude with a Loess smoother having a half span of 10 degrees. Courtesy of D. Chelton, Oregon State University.
Figure 27. Figure provided by D. Chelton, Oregon State University.
Figure 28. Analysis provided by W. Higgins, NOAA Climate Prediction Center.
Figure 29. Analysis provided by W. Higgins, NOAA Climate Prediction Center.
Figure 30. Map of QuikSCAT 10-m vector wind overlaid on sea surface temperature measurements from a satellite microwave radiometer (TMI). Courtesy of D. Chelton.
Figure 31. As in Fig. 14.
Figure 32. US series obtained from the National Climate Data Center Climate Division dataset, and Brazil and Ecuador time series from the NCAR World Monthly Surface Station Climatology.
Figure 33. Analysis provided by W. Higgins, NOAA Climate Prediction Center.
Figure 34. Precipitation time series as in Fig. 32. Sea surface temperature for 1950-1991 from Smith et al. (1996) and for 1992 from Reynolds and Smith (1994). Reynolds, R.W., and T.M. Smith, 1994. Improved global sea surface temperature analyses using optimum interpolation. J. Climate, 7, 929-948. Smith, T. M., R. W. Reynolds, R. E. Livesey, and D. C. Stokes, 1996: reconstruction of historical sea sruface temperatures using empirical orthogonal functions. J. Climate, 9, 1403-1420. As in Fig. 24.
Figure 35. Global tropical cyclone data produced by the National Climatic Data Center, and available online at the NOAA National Hurricane Center and at NCAR.
Figure 36. As in Fig. 35.
No note for Figure 37.
No note for Figure 38.
No note for Figure 39.
No note for Figure 40.