1. Stochastic Hydrology Storm rainfall modeling Professor Ke-Sheng Cheng Department of Bioenvironmental Systems Engineering National Taiwan University

2. Introduction Design storms are routinely used for designing stormwater management facilities and delineating floodplains. A design storm is a hypothetical storm with specific duration D and return period T. The information of a design storms is conveniently presented in the form of depth- duration-frequency (DDF) curves or intensity-duration-frequency (IDF) curves. 11/4/2012 2 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

3. Many engineering designs also rely on the hyetographs, i.e. the time distribution of design storm point rainfall, for runoff calculation. The time distribution of rainfall within a storm event has a significant effect on the peak runoff. Although the shapes of storm hyetographs vary significantly, many studies have shown that dimensionless hyetographs are storm- type specific. 11/4/2012 3 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

4. In contrast to earlier hyetograph models that are duration-specific (Keifer and Chu, 1957; Pilgrim and Cordery, 1975; Yen and Chow, 1980; SCS, 1986), Koutsoyiannis and Foufoula-Georgiou (1993) presented evidence that dimensionless hyetographs are scale invariant. 11/4/2012 4 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

5. Hyetographs from Different Storm Types Huff (1967) normalized the incremental rainfall rates with respect to event-total-depths and storm durations. He classified the storms into four groups, depending on whether the heaviest rainfall occurred in the first, second, third, or the fourth quarter of the storm duration. Huff found that there is a trend for longer, heavier storms to dominate the fourth-quartile group whereas short-duration storms account for a major portion of the first and second quartile groups. 11/4/2012 5 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

6. Eagleson (1970) pointed out that for given climatic conditions, storm events of a given scale (microscale, mesoscale, or synoptic scale) exhibit similar time distributions when normalized with respect to total rainfall depths and storm durations. Convective cells and thunderstorms are dominant types of storms at the microscale and the mesoscale, respectively. Events of synoptic scale include frontal systems and cyclones that are typically several hundred miles in extent and often have series of mesoscale subsystems. In general, convective and frontal-type storms tend to have their peak rainfall rates near the beginning of the rainfall processes, while cyclonic events have the peak rainfall somewhere in the central third of the storm duration. 11/4/2012 6 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

7. Pilgrim and Cordery (1975) developed a hyetograph model based on the average rainfall percentages of ranked rainfalls and the average rank of each time interval in a storm. Koutsoyiannis and Foufoula-Georgiou proposed a simple scaling model to characterize the time distribution of instantaneous rainfall intensity and incremental rainfall depth within a storm event. 11/4/2012 7 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

8. Scaling (Scale Invariant) Strict Sense Simple Scaling Wide Sense Simple Scaling (Second- Order Simple Scaling) Multiple Scaling 11/4/2012 8 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

9. Simple Scaling Model for Storm Events A natural process fulfills the simple scaling property if the underlying probability distribution of some physical measurements at one scale is identical to the distribution at another scale, multiplied by a factor that is a power function of the ratio of the two scales (Gupta and Waymire, 1991). 11/4/2012 9 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

10. Let X(t) and X( t) denote measurements at two distinct time or spatial scales t and t, respectively. We say that the process {X(t), t  0} has the simple scaling property if there is some real number H such that for every real > 0. The denotes equality in distribution, and H is called the scaling exponent. 11/4/2012 10 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

11. Storm events have different durations which may range from a few hours to several days. Traditionally, design storm hyetographs of various storm durations have been developed. For example, the 6-hr and 24-hr SCS hyetographs are commonly used in the practical engineering design. However, if the rainfall process exhibits the simple scaling property, then a scale-invariant hyetograph is desirable. Since the parameter of storm duration does not appear in the preceding simple scaling relation, it is obvious that it cannot be used for translating data between storms of different durations. 11/4/2012 11 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

12. A simple scaling storm model ( Koutsoyiannis and Foufoula-Georgiou, 1993 ) 11/4/2012 12 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

13. 11/4/2012 13 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

14. 11/4/2012 14 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

15. IDF Curves and the Scaling Property The event-average rainfall intensity of a design storm with duration D and return period T can be represented by From the scaling property of total rainfall 11/4/2012 15 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

16. IDF Curves and Random Variables is a random variable and represents the total depth of a storm with duration D. is the (1-p) th quantile (p =1/T) of the random variable, i.e., 11/4/2012 16 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

17. Random Variable Interpretation of IDF Curves 11/4/2012 17 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

18. IDF Curves and the Scaling Property Horner’s Equation: D >> b, particularly for long-duration events. Neglecting b C = - H 11/4/2012 18 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

19. 11/4/2012 19 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

20. Previous studies on the time distribution of rainfall data were based on rainfall data from: – only certain months or a single season (for example, Koutsoyiannis and Foufoula-Georgiou (1993)); or – the entire year (for example, Huff (1967) and Garcia-Guzman and Aranda-Oliver (1993)). Restricting attention to rainfall data in certain seasons allows one to focus on specific storm types and gain better understanding of the dominant and generic storms, as opposed to relying on a hyetograph created by merging rainfall from all storms within a year. 11/4/2012 20 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

21. Because the design storm hyetograph represents the time distribution of the total storm depth determined by annual maximum rainfall data, the design storm hyetograph is optimally modeled when based on observed storm events that actually produced the annual maximum rainfall. Therefore, it is desirable to select only observed storms that give rise to annual maximum rainfalls, the so-called annual maximum events, for development of design hyetographs. 11/4/2012 21 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

22. Annual maximum events Annual maximum events tend to occur in certain periods of the year (such as a few months or a season) and tend to emerge from the same storm type. Moreover, annual maximum rainfall data in Taiwan strongly indicate that a single annual maximum event often is responsible for the annual maximum rainfall depths of different design durations. In some situations, single annual maximum event even produced annual maximum rainfalls for many nearby raingauge stations. 11/4/2012 22 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

23. Examples of annual max events in Taiwan Design durations 11/4/2012 23 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

24. 11/4/2012 24 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

25. Major storm types in Taiwan Storm TypeMei-Yu Convective Storm Typhoon (Cyclonic Storm) Frontal Rainfall Period of Occurrence May - JuneJuly - October November - April 11/4/2012 25 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

26. Statistics for selected annual max events at two sites in Taiwan 11/4/2012 26 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

27. Using rainfall data of the annual maximum events enables us not only to focus on events of the same dominant storm type, it also has the advantage of relying on almost the same annual maximum events that are employed to construct IDF curves. 11/4/2012 27 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

28. Gauss-Markov Model of Dimensionless Hyetographs 11/4/2012 28 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

29. 11/4/2012 29 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

30. 11/4/2012 30 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

31. 11/4/2012 31 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

32. Because the peak rainfall depth is a key element in hydrologic design, an ideal hyetograph should not only describe the random nature of the rainfall process but also the extreme characteristics of the peak rainfall. Therefore, our objective is to find the incremental dimensionless hyetograph that not only represents the peak rainfall characteristics but also has the maximum likelihood of occurrence. 11/4/2012 32 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

33. Our approach to achieve this objective includes two steps: – determine the peak rainfall rate of the dimensionless hyetograph and its time of occurrence, and – find the most likely realization of the normalized rainfall process with the given peak characteristics. 11/4/2012 33 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

34. 11/4/2012 34 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

35. The value of t * is likely to be non-integer and should be rounded to the nearest integer. 11/4/2012 35 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

36. 11/4/2012 36 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

37. Conditional normal density 11/4/2012 37 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

38. 11/4/2012 38 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

39. 11/4/2012 39 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

40. 11/4/2012 40 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

41. 11/4/2012 41 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

42. 11/4/2012 42 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

43. 11/4/2012 43 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

44. 11/4/2012 44 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

45. 11/4/2012 45 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

46. 11/4/2012 46 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

47. The dimensionless hyetograph {y i, i = 1,2, …, n} is determined by solving the matrix equation. 11/4/2012 47 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

48. Model Application Scale-invariant Gauss-Markov model Two raingauge stations Hosoliau and Wutuh, located in Northern Taiwan. 11/4/2012 48 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

49. Annual maximum events that produced annual maximum rainfall depths of 6, 12, 18, 24, 48, and 72-hour design durations were collected. All event durations were first divided into twenty-four equal periods i  (i=1,2, …,24, D = event duration,  =D/24). Rainfalls of each annual maximum event were normalized, with respect to the total rainfall depth and event duration. 11/4/2012 49 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

50. 11/4/2012 50 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

51. Parameters for the distributions of normalized rainfalls 11/4/2012 51 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

52. 11/4/2012 52 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

53. Normality check for normalized rainfalls The Gauss-Markov model of dimensionless hyetographs considers the normalized rainfalls {Y(i), i=1,2, …, n} as a multivariate normal distribution. Results of the Kolmogorov-Smirnov test indicate that at  = 0.05 significance level, the null hypothesis was not rejected for most of Y(i) ’ s. The few rejected normalized rainfalls occur in the beginning or near the end of an event, and have less rainfall rates. 11/4/2012 53 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

54. 11/4/2012 54 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

55. Evidence of Nonstationarity In general, Autocovariance function of a stationary process: For a non-stationary process, the autocovariance function is NOT independent of t. 11/4/2012 55 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

56. 11/4/2012 56 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

57. Calculation of Autocorrelation Coefficients of a Nonstationary Process The lag-k correlation coefficients = correl.(Y(i), Y(i-k)) of the normalized rainfalls were estimated by 11/4/2012 57 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

58. Calculation of Autocorrelation Coefficients of a Nonstationary Process 11/4/2012 58 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

59. 11/4/2012 59 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

60. 11/4/2012 60 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

61. Design storm hyetographs 11/4/2012 61 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

62. 11/4/2012 62 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

63. Translating hyetographs between storms of different durations Some hyetograph models in the literature are duration-specific (the SCS 6-hr and 24- hr duration hyetographs) and return- period-specific (the alternating block method (Chow, et al., 1988). The simple scaling property enables us to translate the dimensionless hyetographs between design storms of different durations. 11/4/2012 63 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

64. Translating the dimensionless hyetographs between design storms of durations D and D is accomplished by changing the incremental time intervals by the duration ratio. Values of the normalized rainfalls Y(i) (i=1,2, …,n) remain unchanged. 11/4/2012 64 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

65. For example, the incremental time intervals (  ) of the design storms of 2-hr and 24-hr durations are five (120/24) and sixty (1440/24) minutes, respectively. The changes in the incremental time intervals are important since they require the subsequent rainfall-runoff modeling to be performed based on the “ designated ” incremental time intervals. 11/4/2012 65 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.

66. A significant advantage of the simple scaling model is that developing two separate dimensionless hyetographs for design storms of 2-hr and 24-hr durations can be avoided. The incremental rainfall depths of a design storm are calculated by multiplying the y-coordinates of the dimensionless hyetograph by the total depth from IDF or DDF curves. 11/4/2012 66 Laboratory for Remote Sensing Hydrology and Spatial Modeling, Dept of Bioenvironmental Systems Engineering, National Taiwan Univ.