MELODIST allows one to derive hourly time series from daily minimum and maximum temperature recordings as depicted below (e.g., Waichler and Wigmosta, 2003):
Besides temperature, sub – daily time series of precipitation are important for many applications in Alpine Hydroclimatology. For instance, water balance simulations for small catchments subjected to short concentration times require adequate precipitation time series.
The figure on the left hand side shows one possible way how to derive hourly precipitation intensities out of daily totals. The method applied is referred to as “cascade” model as it utilises a consecutive doubling of temporal resolution until the desired refinement has been reached (see, Olsson, 1998, Güntner et al., 2001).
To perform this doubling of resolution, comprehensive statistical analyses of hourly time series are required first in order to derive adequate parameters for the model. This step is crucial for adequately disaggregating daily rainfall totals.
Then, these statistical evaluations serve as a base for the disaggregation procedure. The example shows how observed (blue bars) precipitation is refined from 24 h to 1 h temporal resolution. The green bars represent the disaggregated time series for each step:
24 h, 12 h, 6 h, 3 h, 1.5 h, 45 min (not shown), and 1 h
Each run of the model generates different results due to the application of a random number generator. However, these results are similar with respect to statistical characteristics. Hence, another run using the data from the example would yield a different sub – daily redistribution of precipitation time series (i.e., rainfall in the morning with smaller intensity might be possible). Therefore, this stochastic type model is valid for long – term evaluations covering several years rather than for single events as exemplary shown here.
A long term evaluations of precipitation time series is presented in a discussion paper which is in review at present(Förster et al., 2016).MELODIST is also capable of disaggregating humidity, radiation, and wind speed based on methods described by Liston and Elder 2006, Debele et al.(2007), and Bregaglio et al.(2010).All variables are treated independently.A detailed description of the methodology along with a validation of the methods based on stations recordings in different climates is described by Förster et al., (2016).
The software is available here:
Bregaglio, S., Donatelli, M., Confalonieri, R., Acutis, M., and Orlandini, S. (2010): An integrated evaluation of thirteen modelling solutions for the generation of hourly values of air relative humidity. Theor. Appl. Clim., 102 (3-4), 429-438.
Debele, B., Srinivasan, R., and Parlange, J. Y. (2007): Accuracy evaluation of weather data generation and disaggregation methods at finer timescales. Adv. in Water Res., 30 (5), 1286-1300.
Förster, K., Hanzer, F., Winter, B., Marke, T. and Strasser, U. (2016): An open-source MEteoroLOgical observation time series DISaggregation Tool (MELODIST v0.1.1), Geosci. Model Dev., 9, 2315-2333, doi:10.5194/gmd-9-2315-2016.
Güntner, A., Olsson, J., Calver, A., and Gannon, B. (2001): Cascade-based disaggregation of continuous rainfall time series: the influence of climate, Hydrol. Earth Syst. Sci., 5, 145-164, doi:10.5194/hess-5-145-2001.
Liston, G. E. and Elder, K. (2006): A meteorological distribution system for high-resolution terrestrial modeling (MicroMet). J. Hydrometeorol., 7 (2), 217-234.
Olsson, J. (1998): Evaluation of a scaling cascade model for temporal rain- fall disaggregation, Hydrol. Earth Syst. Sci., 2, 19-30, doi:10.5194/hess-2-19-1998.
Waichler, S. R. and Wigmosta, M. S. (2003): Development of hourly meteorological values from daily data and significance to hydrological modeling at HJ Andrews Experimental Forest. J Hydrometeorol., 4 (2), 251-263.