On the Choice of the Temporal Aggregation Level for Statistical Downscaling of Precipitation

T. A. Buishand Royal Netherlands Meteorological Institute, De Bilt, Netherlands

Search for other papers by T. A. Buishand in
Current site
Google Scholar
PubMed
Close
,
M. V. Shabalova Royal Netherlands Meteorological Institute, De Bilt, Netherlands

Search for other papers by M. V. Shabalova in
Current site
Google Scholar
PubMed
Close
, and
T. Brandsma Royal Netherlands Meteorological Institute, De Bilt, Netherlands

Search for other papers by T. Brandsma in
Current site
Google Scholar
PubMed
Close
Restricted access

Abstract

The merits of daily and monthly downscaling models for precipitation are compared using data from Bern, Switzerland; Deuselbach, Germany; and De Bilt, the Netherlands. For each station, generalized linear models are developed to describe rainfall occurrence, the wet-day precipitation amounts, and the monthly precipitation totals. The predictor dataset includes dynamical variables and atmospheric moisture (relative humidity for rainfall occurrence and specific humidity for rainfall amount).

Fitting a generalized linear model to daily rainfall data generally results in larger regression coefficients than fitting the same model to monthly data. For rainfall occurrence this can be attributed mostly to the nonlinearity of the function that links the wet-day probabilities to the predictor variables, whereas for rainfall amounts there is, apart from nonlinearity, also a bias in the estimated regression coefficients of the monthly models caused by averaging predictor variables over both wet and dry days. Because of this bias a monthly rainfall amount model is less sensitive to an increase in the specific humidity than a daily rainfall amount model.

Although the squared correlation coefficient r2 between the observed and predicted values of the daily models is low (≈0.40 for rainfall occurrence and ≈0.15 for wet-day rainfall), aggregating the results from these models to monthly values gives r2 values comparable to those in the direct fit to the monthly data (≈0.65 for the number of wet days and ≈0.50 for rainfall totals). The temporal variations in the predicted annual amounts using monthly relationships are similar to those obtained from daily relationships. Daily models are preferable, however, for the generation of climate change scenarios for impact studies, because the significance of the predictor variables is generally stronger in these models and because the effect of a change in specific humidity is underestimated by the monthly models.

Corresponding author address: Dr. T. A. Buishand, Royal Netherlands Meteorological Institute (KNMI), P.O. Box 201, 3730 AE De Bilt, Netherlands. Email: Adri.Buishand@knmi.nl

Abstract

The merits of daily and monthly downscaling models for precipitation are compared using data from Bern, Switzerland; Deuselbach, Germany; and De Bilt, the Netherlands. For each station, generalized linear models are developed to describe rainfall occurrence, the wet-day precipitation amounts, and the monthly precipitation totals. The predictor dataset includes dynamical variables and atmospheric moisture (relative humidity for rainfall occurrence and specific humidity for rainfall amount).

Fitting a generalized linear model to daily rainfall data generally results in larger regression coefficients than fitting the same model to monthly data. For rainfall occurrence this can be attributed mostly to the nonlinearity of the function that links the wet-day probabilities to the predictor variables, whereas for rainfall amounts there is, apart from nonlinearity, also a bias in the estimated regression coefficients of the monthly models caused by averaging predictor variables over both wet and dry days. Because of this bias a monthly rainfall amount model is less sensitive to an increase in the specific humidity than a daily rainfall amount model.

Although the squared correlation coefficient r2 between the observed and predicted values of the daily models is low (≈0.40 for rainfall occurrence and ≈0.15 for wet-day rainfall), aggregating the results from these models to monthly values gives r2 values comparable to those in the direct fit to the monthly data (≈0.65 for the number of wet days and ≈0.50 for rainfall totals). The temporal variations in the predicted annual amounts using monthly relationships are similar to those obtained from daily relationships. Daily models are preferable, however, for the generation of climate change scenarios for impact studies, because the significance of the predictor variables is generally stronger in these models and because the effect of a change in specific humidity is underestimated by the monthly models.

Corresponding author address: Dr. T. A. Buishand, Royal Netherlands Meteorological Institute (KNMI), P.O. Box 201, 3730 AE De Bilt, Netherlands. Email: Adri.Buishand@knmi.nl

Save
  • Beckmann, B-R., and T. A. Buishand, 2002: Statistical downscaling relationships for precipitation in the Netherlands and north Germany. Int. J. Climatol, 22 , 1532.

    • Search Google Scholar
    • Export Citation
  • Buishand, T. A., 1978: Some remarks on the use of daily rainfall models. J. Hydrol, 36 , 295308.

  • Buishand, T. A., and B-R. Beckmann, 2000: Development of daily precipitation scenarios at KNMI. Climate Scenarios for Water-Related and Coastal Impacts, J. Beersma et al., Eds., ECLAT-2 Workshop Rep. 3, Climatic Research Unit, University of East Anglia, 79– 91. [Available from Climatic Research Unit, University of East Anglia, Norwich NR4 7TJ, United Kingdom.].

    • Search Google Scholar
    • Export Citation
  • Buishand, T. A., and A. M. G. Klein Tank, 1996: Regression model for generating time series of daily precipitation amounts for climate change impact studies. Stochastic Hydrol. Hydraul, 10 , 87106.

    • Search Google Scholar
    • Export Citation
  • Coe, R., and R. D. Stern, 1982: Fitting models to daily rainfall data. J. Appl. Meteor, 21 , 10241031.

  • Frei, C., and C. Schär, 2001: Detection probability of trends in rare events: Theory and application to heavy precipitation in the Alpine region. J. Climate, 14 , 15681584.

    • Search Google Scholar
    • Export Citation
  • Giorgi, F., and Coauthors, 2001: Regional climate information—Evaluation and projections. Climate Change 2001: The Scientific Basis, J. T. Houghton et al., Eds., Cambridge University Press, 583– 638.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc, 77 , 437471.

  • Katz, R. W., 1977: Precipitation as a chain-dependent process. J. Appl. Meteor, 16 , 671676.

  • Kendall, M., A. Stuart, and J. K. Ord, 1983: The Advanced Theory of Statistics. Vol. 3. 4th ed. Charles Griffin, 780 pp.

  • Kilsby, C. G., P. S. P. Cowpertwait, P. E. O'Connell, and P. D. Jones, 1998: Predicting rainfall statistics in England and Wales using atmospheric circulation variables. Int. J. Climatol, 18 , 523539.

    • Search Google Scholar
    • Export Citation
  • Kistler, R., and Coauthors, 2001: The NCEP–NCAR 50-year reanalysis: Monthly means CD-ROM and documentation. Bull. Amer. Meteor. Soc, 82 , 247267.

    • Search Google Scholar
    • Export Citation
  • Lemcke, C., and S. Kruizinga, 1988: Model output statistics forecasts: Three years of operational experience in the Netherlands. Mon. Wea. Rev, 116 , 10771090.

    • Search Google Scholar
    • Export Citation
  • McCullagh, P., and J. A. Nelder, 1989: Generalized Linear Models. 2d ed. Chapman and Hall, 511 pp.

  • Merenti-Välimäki, H-L., and P. Laininen, 2002: Analysing effects of meteorological variables on weather codes by logistic regression. Meteor. Appl, 9 , 191197.

    • Search Google Scholar
    • Export Citation
  • Mitchell, J. F. B., and W. J. Ingram, 1992: Carbon dioxide and climate: Mechanisms of changes in cloud. J. Climate, 5 , 521.

  • Murphy, J., 1999: An evaluation of statistical and dynamical techniques for downscaling local climate. J. Climate, 12 , 22562284.

  • Murphy, J., 2000: Predictions of climate change over Europe using statistical and dynamical downscaling techniques. Int. J. Climatol, . 20 , 489501.

    • Search Google Scholar
    • Export Citation
  • Reid, P. A., P. D. Jones, O. Brown, C. M. Goodess, and T. D. Davies, 2001: Assessments of the reliability of NCEP circulation data and relationships with surface climate by direct comparisons with station based data. Climate Res, 17 , 247261.

    • Search Google Scholar
    • Export Citation
  • Semenov, V. A., and L. Bengtsson, 2002: Secular trends in daily precipitation characteristics: Greenhouse gas simulation with a coupled AOGCM. Climate Dyn, 19 , 123140.

    • Search Google Scholar
    • Export Citation
  • Stern, R. D., and R. Coe, 1984: A model fitting analysis of daily rainfall data (with discussion). J. Roy. Stat. Soc, 147A (1) 134.

  • Weisberg, S., 1985: Applied Linear Regression. 2d ed. John Wiley and Sons, 324 pp.

  • Wilby, R. L., 2001: Downscaling summer rainfall in the UK from North Atlantic Ocean temperatures. Hydrol. Earth Syst. Sci, 5 , 245257.

    • Search Google Scholar
    • Export Citation
  • Wilby, R. L., H. Hassan, and K. Hanaki, 1998: Statistical downscaling of hydrometeorological variables using general circulation model output. J. Hydrol, 205 , 119.

    • Search Google Scholar
    • Export Citation
  • Wilby, R. L., C. W. Dawson, and E. M. Barrow, 2002: SDSM—A decision support tool for the assessment of regional climate change impacts. Environ. Modell. Software, 17 , 147159.

    • Search Google Scholar
    • Export Citation
All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 300 59 2
PDF Downloads 106 39 2