1
Article
5
Development of Monsoonal Rainfall IntensityDurationFrequency (IDF) Relationship and Empirical Model for datascarce situations: the case of CentralWestern Hills (Panchase region) of Nepal
6 7
Sanjaya Devkota1, Narendra Man Shakya1, Karen SudmeierRieux2, Michel Jaboyedoff2, Cees J. Van Westen3, Brian G. Mcadoo4, Anu Adhikari5
2 3 4
8 9 10 11 12 13
1
14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
Abstract: Intense monsoonal rain is one of the major triggering factors of floods and mass movements in Nepal that needs to be better understood in order to reduce human and economic losses and improve infrastructure planning and design. This phenomena is better understood through intensitydurationfrequency (IDF) relationships, which is a statistical method derived from historical rainfall data. In Nepal the use of IDF for disaster management and project design is very limited. This study explored the rainfall variability and possibility to establish IDF relationships in data scarce situations, such as in the CentralWestern hills of Nepal, one of the highest rainfall zones of the country (~4,500mm annually), which was chosen for this study. Homogeneous daily rainfall series of 8 stations, available from the meteorological government department were analyzed grouping them into hydrological years. The monsoonal daily rainfall was disaggregated to hourly synthetic series in a stochastic environment. Utilizing the historical statistical characteristics of rainfall, a disaggregation model was parameterized and implemented in HyetosMinute, software that disaggregates daily rainfall to finer time resolution. With the help of recorded daily and disaggregated hourly rainfall, reference IDF scenarios were developed adopting Gumbel frequency factor. A mathematical model [i=a(T)/b(d)] was parameterized to model the stationspecific IDF utilizing the best fitted PDF and evaluated utilizing the reference IDF. The test statistics revealed optimal adjustment of empirical IDF parameters, required for better statistical fit of the data. The model was calibrated, adjusting the parameters by minimizing standard error of prediction; accordingly a stationspecific empirical IDF model was developed. To regionalize the IDF for ungauged locations, regional frequency analysis (RFA) based on Lmoments was implemented. The heterogeneous region was divided into two homogeneous subregions accordingly regional Lmoment ratios and growth curves were evaluated. Utilizing the reasonably acceptable distribution function regional growth curve was developed. Together with the hourly mean (extreme) precipitation and other dynamic parameters regional empirical IDF models were developed. The adopted approach to derive station specific and regional empirical IDF model were statistically significant and useful to obtain extreme rainfall intensities at the given station and ungauged locations. The analysis revealed that the region contains two distinct meteorological subregions with highly variable in rain volume and intensity.
42 43
Keywords: Nepal; monsoonal rain; data scarcity; IntensityDurationFrequency, Lmoment
Trivbhuvan University, Department of Civil Engineering, Institute of Engineering, Lalitpur, Nepal University of Lausanne, Institute of Earth Science (ISTE), Lausanne, Switzerland 3University of Twente, Faculty of Geoinformation Science and Earth Observation, the Netherlands 4 YaleNUS College, Singapore 5International Union for Conservation of Nature (IUCN), Lalitpur, Nepal Corresponding Author: Sanjaya Devkota, Email:
[email protected] 2
Hydrology 2018, 5, x; doi: FOR PEER REVIEW
www.mdpi.com/journal/hydrology
Hydrology 2018, 5, x FOR PEER REVIEW
2 of 31
44
1. Introduction
45
1.1. Background
46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85
Nepal has experienced numerous natural hazard events over the past resulting in enormous economic losses and thousands of causalities, compounding the country’s already high level of poverty [15]. The frequent natural and humaninduced disasters are due to its fragile geomorphology, active tectonics, and unplanned human activity on steep slopes combined with climate extremes [3, 4, 68]. Flooding and landslides occur frequently, mostly triggered by extreme precipitation or by earthquakes, as evidenced by the 2015 Gorkha earthquake. The importance of human interventions, such as road construction, on mass movements was at best underestimated and largely neglected by researchers and authorities in Nepal [9]. Landslides are caused by a number of underlying natural and human factors including slope aspect, gradient, soil type, land cover (changes), proximity to rivers and land use (i.e. road construction, quarrying) and most often triggered by rainfall or earthquakes [8, 10]. Rainfall, while critical to maintaining ecosystem services and supporting livelihoods, is thus a triggering factor that needs to be understood and managed in order to reduce impacts due to flooding and mass movements and erosion. Reducing and managing climateinduced disasterrisk (e. g. damage due to floods and landslides) relies on knowledge of the frequency and intensity of rainfall events [11]. Extreme precipitation induced pluvial flood and shallow landslides disaster management requires the establishment of intensitydurationfrequency (IDF) relationships of extreme events, to formulate better design guidelines for the development of infrastructure in order to reduce impacts due to disasterrisk, and save lives, properties and ecosystems [12]. However, inadequate rainfall data continue to hamper the establishment of reasonable IDF relationships [1315]. Researchers have realized that the design of water resources projects, management of storm water runoff and disaster mitigation planning with scarce and insufficient data is always a challenge [13, 16, 17]. The situation is even more challenging in least developed countries such as Nepal. In Nepal, the Department of Hydrology and Meteorology (DHM) is the government organization responsible for meteorological instrumentation and maintaining the climate variable database including rainfall. DHM has reported that there is no instrumentation for measuring finescale rainfall (e. g. hourly, subhourly precipitation) nor did it develop any systematic IDF relationships for the country. This research explored the possibility to establish an IDF relationship in the data scarce environment of Nepal and developed an empirical IDF model to better estimate extreme rainfall events at some fixed duration of time and recurrence interval, for the Panchase region in the CentralWestern Hills of Nepal, a region with one of the highest annual rainfall amounts (between 4,0005,000 mm, mean = 2,984.40mm, SD = 1,497.60). The objectives of the study were to develop monsoon season IDF relationships for the low resolution rainfall data recording situation followed by the development of location specific empirical IDF models and demonstrate the application of IDF for ungauged locations. For this purpose, the recorded daily and disaggregated hourly rainfall series were the main data sources. Using the IDF, we can better understand the short and long term rainfall intensity during the monsoon season that triggers mass movements and other hazards induced by intense rainfall in this region. Further, the IDF relationship can be used in water resources management and infrastructure planning and design.
86
1.2. Rational
87 88 89 90 91 92 93
The importance of rainfall and IDF relationships to address extreme precipitation induced hazards such as massmovement have emphasized by many researchers [1820]. IDF curves are also the graphical representation that summarizes the important statistical properties of extreme precipitation events [21]. First established in 1932 by Bernard [17], IDF is a statistical relationship of the intensity, duration and frequency of rainfall derived from historical rainfall data [13, 16, 22]. Since then, many different sets of relationships have been constructed for different parts of the world [13, 16, 17, 2225].
Hydrology 2018, 5, x FOR PEER REVIEW
3 of 31
94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
In the literature there are several functions to establish IDF relationships (e. g. [16, 22, 26]). Chen [26] derived a generalized IDF relationship for any location in the United States using three basic rainfall depths (e. g. 1 hour 10year, 24 hour 10year, and 1 hour 100year rainfall depth). Baghirathan and Shaw [27] and Gert et al. [28] proposed regional IDF formulae for ungauged areas while Kothyari and Garde [29] used daily rain and two year return period to establish the IDF relation in India. Further, Koutsoyiannis et al. [22] proposed a mathematical approach to formulate IDF relationships followed by the construction of IDF curves utilizing point information of long term historical rainfall time series in the context of geographical variability and regionalization of IDF model. The concept of regional IDF relationships were executed by Yu and Chen [30] and Madsen et al. [31] who examined regression techniques while Willems [32], Yu et al. [33] and Langousis and Veneziano [19] applied scaling method and developed regional IDF curves. Dalrymple [34] proposed regional frequency analysis (RFA) method for pooling various data samples also known as indexflood procedure in hydrology [35]. Hosking et al. [36] studied the properties of probabilityweighted moments (PWMs) method based on Lmoment and Hosking and Wallis [11] showed the application Lmoments for the RFA. Lmoments are efficient tool used to detect the homogeneous regions, to select suitable regional frequency distribution, and to predict extreme precipitation quantiles at region of interest. IDF relationship if regionalized can minimize computational time and effort to obtain the IDF curves for areas where rainfall gauging stations are not installed. In order to establish IDF relationships in data scarce situations, researchers need to disaggregate commonly available daily rainfall data. They have therefore developed methods for utilizing historical statistical information from commonly available daily rainfall to synthetically generate fine time resolution rainfall series [13, 3744]. Stochastic simulation tools generate fine timescale synthetic rainfall series from coarser resolution preserving similar statistical properties [37, 3941, 45]. Nepal as a whole is dominated by SE monsoon [46], where topography has a considerable effect on the rainfall patterns. There are four distinct hydrological seasons: premonsoon (April and May: AP), monsoon (June to September: JJAS), postmonsoon (October and November: ON) and winter or dry period (December to March: DJFM) in the country [46]. The monsoonal rainfall is highly variable over the country (~1,000mm~4,500mm) and is intense in nature where more than 80% of the annual rainfall occurs in the 4 months of the monsoon [18, 47], resulting in landslides, debris flows, flooding and sedimentation, threatening to the livelihoods and properties in numerous ways [48].
128
1.3. The Panchase Region
129 130 131 132 133 134 135 136
Panchase is a mountainous region in the middlehills of CentralWestern Nepal between latitudes 280 12’N to 280 18’ N, and longitudes 830 45’ E to 830 57’ E. The region is located in three districts – Kaski, Syangja and Parbat (Fig. 1). The elevation varies from 742 meter above sea level (masl) (outlet of Phewa Lake, near Pokhara city) to 2,523 masl (Panchase Peak) and characterized by hot, humid summers and cooltemperate winter seasons [6]. The Panchase hill range is extending from southwest to northwest direction dividing the region into two distinct eastern and western regions. In order to evaluate the rainfall variability and to establish IDF relationship of the region the historical rainfall data of 11 weathers station were collected from the DHM (Table 1).
137 138
Table 1: Summery statistics of historical daily rainfall (19822015), missing values and homogeneity test of all 11 stations. DHM
Annual
Monsoonal
Nr. of
Nr. of Homogeneity
Station
Altitude
Years of
(mean)
(mean)
storms
storms
Missing
Number
(masl)
Records
rainfall
rainfall
exceeding
exceeding
Values
(mm)
(mm)
100 mm in
200 mm in
Location
(rejected at 95%) (Nr.)
Hydrology 2018, 5, x FOR PEER REVIEW
4 of 31
24 hrs
24 hrs
GHANDRUK
821
1960
19822014
3384
2642
47
0
>5%
Yes
LUMLE
814
1740
19822014
5504
4681
352
17
< 5%
No
KARKINETA
613
1720
19822014
2543
2047
56
0
< 5%
No
813
1600
19842015
3744
3093
150
7
< 5%
No
LAMACHAUR
818
1070
19822014
4220
3380
204
10
>5%
Yes
KUSHMA
614
891
19822014
2531
2122
39
0
< 5%
No
SYANGJA
805
868
19822014
2840
2280
101
7
< 5%
No
804
827
19822015
3969
3160
189
19
< 5%
No
BHADAUREDEURALI
POKHARA AIRPORT WALLING
826
750
19892012
1929
1658
61
6
< 5%
No
KHAIRINITAR
815
500
19822012
2384
1719
50
1
< 5%
No
CHAPAKOT
810
460
19822012
1878
1451
59
2
>5%
Yes
139 140 141 142 143 144 145 146 147 148
The region is one of the most studied areas in the country [4853] due to the importance of Phewa Lake that promotes economic activities and biodiversity. However, the study of rainfall patterns, its intensity and impacts have not been properly addressed yet. The Ecosystem Protecting Infrastructures and Communities (EPIC) project established and monitored three communitybased rural roadside bioengineering demonstration sites in the region. In addition to capacity building and policy advocacy, the project combined formal and citizenscience by mobilizing local people and exploring local knowledge regarding climate extremes, plant species and the importance of rural access roads for building climate resilient communities. In order to link local knowledge with formal science, the project installed three continuous recording meteorological stations in the period NovemberFebruary 2014 next to each demonstration site.
149 150 151
Fig. 1 Study Region indicating the location (name and station number) of weather stations of the DHM and those installed from the EPIC project (Table 1 for summary).
152
2. Methods
153
2.1. Data Quality and Rainfall Variability
154 155 156 157 158 159
Historical daily rainfall data of 11 weather stations (Table 1 and Fig. 1) in and around Panchase region with a recording period of above 30 years were obtained from DHM. Only the station in Walling had a shorter recording period of 22 years. Frequency analysis of rainfall time series requires that the data are homogeneous and independent [54]. In order to evaluate the data quality we performed homogeneity test and evaluated for missing values. Among several methods of homogeneity test (see [55]) we implemented simple cumulative deviations of the data from the mean
Hydrology 2018, 5, x FOR PEER REVIEW
160 161
5 of 31
according to Buishand [54] to check whether the sample data were from the same population (Equation 13, Table 1). The cumulative deviations from the mean can be expressed as:
k S X X k i i 1 162
where,
163 164 165 166 167
Sk 0
k 1, 2,......., n
(1)
X are the records from the meteorological series X1 , X 2 , …… X n and X is the mean. i and Sk n respectively the initial and final value of the series.
Buishand [54] demonstrated cumulative deviations of the mean of rainfall records are often rescaled dividing
S k by the sample standard deviation ( ). By evaluating the maximum (Q) or the
range (R) of the rescaled cumulative deviations from the mean, the homogeneity of the data series can be tested, where:
Q
Sk
max 0 k n x
S S R max k min k 0 k n x 0 k n x
(2)
(3)
168 169 170 171 172 173 174 175 176 177 178 179 180
According to Raes et al. [56] high values of Q and R indicate that data of the series are not from the same population and the fluctuations are not purely random. The inhomogeneous data series were excluded from the analysis whereas homogeneous data series containing less than 5% missing values were further evaluated. To complete the data series missing values were imputed using the RClimTool [57] and nearest neighborhood method as explained in XLSTAT [58]. To understand the rainfall variability in Panchase region we adopted a generic approach of analyzing the completed daily rainfall series in terms of variation of annual monsoonal rain [JJAS], variation in monsoonal dry days and extremes events of over 100 mm of rain depth in 24 hours. To better estimate the variation we grouped the rainfall series according to the hydrological year that start from April and focused on monsoonal rainfall depth. Evaluation of historical monsoonal dry/wet days can be an indicator to understand the rainfall variability as shown by other researchers (e. g. [59]).
181
2.2. Disaggregation of Daily Rainfall
182 183 184 185 186 187 188 189 190 191 192 193 194
A general framework to generate synthetic rainfall time series at finer time resolution consistent with the given coarser resolution, preserving the statistical characteristics (e.g. mean, standard deviation, lag 1 autocorrelation and percentage of dry days) of both scales, was initially proposed by Koutsoyiannis [42] and Koutsoyiannis and Manetas [60]. Koutsoyiannis and Onof [44] extended the methodology to disaggregate daily rainfall into hourly by coupling with the BartlettLewis Rectangular Pulse (BLRP) Model [41, 44] with adjusting procedure [60]. The original BLRP model proposed by RodriguezIturbe et al. [61] consists of five parameters ( , , , , and x ) and
modelled the rainfall using rectangular pulses and characterized by the parameters (Fig. 2). The general assumption of the BLRP model as proposed by RodriguezIturbe et al. [61] are: The occurrence of random storm events (ti) are assumed to be modelled as a Poisson process with rate and each event i is associated with a random number of cells Each storm events tij, is assumed as precipitation rectangular pulse with random duration td and the origin of storm events tij of each cell j occurs following second Poisson process with
Hydrology 2018, 5, x FOR PEER REVIEW
195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212
213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235
6 of 31
rate . The interarrival time of two subsequent storm events (i. e. successive cells) are independent, identically distributed and follow an exponential distribution
The cell generation process terminates after time span of
i following the exponential
distribution rate . Also the number of cells per storm contains a geometric distribution of mean c 1 .
The random precipitation rectangular pulse duration td is modelled as wij also follows exponential distribution with rate .
Finally, the cell intensity xij is assumed to be exponentially distributed with mean x .
According to RodriguezIturbe et al. and Onof and Wheater [62, 63] the model BLRP reproduced the basic statistics but observed difficulties in reproducing the temporal characteristics. According to Kossieris et al. [41] in order to improve the model’s flexibility in generating a greater diversity of rainfalls, RodriguezIturbe et al. [63] modified the original model so that the parameters is randomly varied from storm to storm according to gamma distribution with a shape parameter
and rate parameter
( ) to the parameter
in such way that the ratios of cell origin rate ( ) and storm duration rate
(i. e. and ) are kept constant. Also, the parameters
are random variables that follow a gamma distribution with common shape parameter and rate parameters and [41]. Implementing the modified approach the model consist of and
six parameters ( , , ,
,
and
x ) what we called modified BLRP (MBLRP).
Fig. 2 Graphical representation of the BLRP model according to [61]. Filled and open circles denote respectively the storm origins and cell arrivals [41].
The successful application of the MBLRP model in different climatic regions was reported by several researchers (e. g. [38, 39, 41, 44, 62, 6471]. The model was successfully verified and implemented through HyetosMinute, and Rbased computer software by Kossieris et al. [41]. According to Kossieris et al. [41], HyetosMinute is an extended and improved version of the Hyetos model [43] to disaggregate daily rainfall depth to subhourly resolution [72]. HyetosMinute implements the MBLRP scheme of the Poissonclusters model and disaggregates daily rainfall, dividing the rainfall depth into many clusters as sequences of wet days separated by at least one dry day. Also, each sequence is considered to be an independent event [39, 41]. We used the HyetosMinute [72] to generate the required level of rainfall resolution according to Kossieris et al. [41]. As far as we are aware this is the first attempt of applying a rainfall disaggregation approach in order to establish IDF relationships in Nepal and thus we evaluated the model parameter for each station first for the annual and second for the seasonal rainfall series. We focused on available daily monsoonal rainfall series to parameterize the model adopting a global optimization algorithm. The algorithm was implemented in computer package called Evolutionary AnnealingSimplex (EAS) in R software [41, 72] utilizing the historical statistical characteristics in terms of mean, variance, covariance, and percentage of dry days. While performing the EAS, suitable boundary conditions of the model parameters (Table 2) were identified for the monsoon season daily time series rainfall executing several iterations, in such a way that the model preserves the statistical characteristics of the historical rainfall.
Hydrology 2018, 5, x FOR PEER REVIEW
236 237 238 239
7 of 31
Table 2: Adopted Lower and Upper bounds to calibrate global optimization algorithms implemented in EAS software package (where = rate of storms occurrence, =shape parameter for gamma distribution, = rate of change of storms,
, and
x = rain
,
intensity) .
Parameter
(day1)
(day)
x (mm/day)
Lower Bound
0.001
0.001
0.001
0.001
0.001
0.001
Upper Bound
10
0.1
20
20
1
60
240 The global optimization algorithm generated statistical parameters ( , , ,
,
x )
241 242 243
derived from the daily available monsoonal rainfall data and rainfall depth as an input disaggregation was executed in HyetosMinute [41, 72].
244
2.3. Evaluation of Disaggregation Model
245 246 247 248 249 250 251 252 253 254 255 256
In the case of Nepal, we are facing an inadequate data situation; hence we are limited in our ability to tune the model. We performed the model evaluation using limited fine resolution (hourly) rainfall data recorded within the study region. We implemented the model performance evaluation scheme according to Kossieris et al. [41]. The EPIC project has established three tippingbucket type weather stations within the test area which were calibrated to measure rainfall volume of 0.2 mm per tip with a temporal resolution of an hour. Hourly rainfall data (JuneSept 2016) of Gharelu, one of the test sites of EPIC were obtained and used to evaluate the disaggregation model performance. For this purpose we first aggregated the hourly rainfall to daily series then the series was disaggregated adopting the procedure discussed in section 2.2. The statistical characteristics of the aggregated daily series was calculated and fitted into the EAS model to estimate the parameters (Table 3). To maintain the consistency in the disaggregation procedural we adopted similar boundary conditions as in Table 2 above.
257 258
Table 3: Estimated MBLRP parameters to be used in HyetosMinute derived from the test rainfall data series.
Location
Weather station
Gharelu
EPIC :1
(day ) 1.4805
1.9181
0.2615
(day)
1
6.9398
0.1534
and
x (mm/day)
105.98
259 260 261 262
We calculated the Mean (En), Variance (Varn) and Skewness (Skewn) of the disaggregated series and compared with the characteristic of aggregated series for each of the monsoonal months. Further, we evaluated the recorded and disaggregated hourly intensity for the test data series for which a Goodness of Fit (GoF) test was performed.
263
2.4. Probability Distribution Function for the Rainfall Data and Goodness of Fitting
264 265 266 267 268 269 270 271
The probability distribution of time series with daily rainfall data is important in the field of hydrology and meteorology [73]. It is also important for the construction of IDF curves that requires the fitting of a Probability Distribution Function (PDF) according to Koutsoyiannis et al. [22]. Several distribution models (e. g. Gumbel: Extreme Value Type I, Generalized Extreme Value: GEV, Log Pearson Type III, Beta, Gamma, lognormal, Normal, etc.) are particularly useful for hydrological and meteorological time series data analysis [22, 7477]. In Nepal there is currently limited use of IDF curves/model, and thus no preferred distribution model to be fitted for the rainfall data series. However, in some flood frequency studies, the Gumbel Extreme Value Type I and LogPearson Type
Hydrology 2018, 5, x FOR PEER REVIEW
272 273 274 275 276 277 278
8 of 31
III were used [52]. In this research, the PDF was evaluated and the best fit was chosen for which EasyFit statistical software developed by MathWaveTechnology [78] was used as used in Gamage et al. [79] and Misic [80] and as discussed in Hosking et al. [81], Stedinger et al. [82], Koutsoyiannis [83, 84] and Millington et al. [75]. We chose the EasyFit because it is robust and capable to handle small to large time series data. The homogeneous daily rainfall data of eight weather stations discussed earlier (see Table 1) were examined by fitting the PDF in particular Generalized Extreme Value (GEV) and Gumbel Extreme Value Type I (EV I) (Equation 45). 1 x Fx ( x) exp 1
279 280
where,
0 , 0 and
0
(4)
are shape, scale and location parameters respectively. For 0 the
GEV distribution turns into the Gumbel distribution [82]:
x Fx ( x) exp exp
(5)
281 282 283 284 285 286
While analyzing the daily time series we recognized that the data contain some degree of seasonality effects, leading to separation of the annual series into seasonal to better understand the rainfall characteristics. The PDF for the monsoonal daily and annual extreme time series rainfall was assessed using the test statistics adopting KolmogorovSmirnov (KS), AndersonDarling (AD) and ChiSquared test ( 2 ) as illustrated in the EasyFit software [78]. Accordingly best fitted PDF was
287
2.5. Development of IDF Relationship and Empirical Model
288 289 290 291 292
In this study we propose to derive a reference IDF utilizing the recorded monsoonal daily and disaggregated hourly extreme rainfall followed by implementation of mathematical model according to Koutsoyiannis et al. [22]. We computed the frequency of precipitation depth PT, (in mm) for the given rainfall duration td (in hour) with specified return period Tr (in Years) according to Wilson [85]:
chosen and model parameters ( , and ) were estimated.
PT Pave KT Sd 293 294 295
PT is frequency of precipitation depth, Pave is mean rainfall of monsoonal annual series (recorded daily and disaggregated hourly) extremes, Sd is standard deviation of annual series and KT is Gumbel frequency factor for the given duration (Equation 7): where,
KT 296 297 298
(6)
Tr 6 0.5772 ln ln Tr 1
(7)
where, Tr is return period and 0.5772 is Euler’s constant. Applying the following relation (Equation 8) we obtained the rainfall intensity IT (mm/h) for the given duration td (1 hr and 24 hr) and return period Tr (2, 5, 10, 25, 50, 100 and 200 years):
IT
PT td
(8)
Hydrology 2018, 5, x FOR PEER REVIEW
299 300 301 302 303
9 of 31
From Equation 6, 7 and 8, we determined the rainfall intensities for given duration and given return period, what we called IDF scenarios, representing given locations (i. e. stations). These IDF scenarios were later used as reference to evaluate stationspecific fine resolution IDF curves developed through implementing the mathematical model proposed by Koutsoyiannis et al. [22] expressed as below:
i 304 305 306 307 308 309 310
a T b( d )
Koutsoyiannis et al. [22] provided that the nominator a(T ) can be determined through the PDF of the data set and the denominator b( d ) is the function of duration ( td ) with some empirical
and . The denominator defines the shape of the curves where the parameter is the indicative of the slope of the curves and characterizes the timing of the change in the curvature parameters
and is dependent on the duration of the precipitation [17]. This relationship of intensity i can be separated into a(T ) and b( d ) for which Koutsoyiannis et al. [22] demonstrated following expressions (Equation 10 and 11), for a(T ) and b( d ) respectively.
1 a(T ) ln ln 1 Tr
311 312
(9)
where, and
are scale and location parameters of the distribution respectively, Tr is return
period for the best fitted PDF (i.e. EV I, see section 3.4).
b(d ) (td )
313 314 315 316 317 318 319 320 321 322 323
(10)
where,
(11)
td is duration, and are parameter to be estimated ( >0 and 0< 3.
376
2.6.2 Identification of Homogeneous Region
377 378 379 380 381 382 383 384 385 386 387
This is the core part of RFA and based on the grouping of sites to regions in which the sample are from the same population and consist of same frequency distributions. We implemented the Lmoment based RFA dividing the eight rain gauge stations distributed into the heterogeneous region to two homogeneous subregions (eastern region: Lumle, Bhadaure, PokharaAirport and Khairenitar; western region: Kusma, KarkiNeta, Syangja and Walling). The division was made considering the meteorological and physical characteristics such as mean annual precipitation, altitude, latitude and distance to the gauging stations to the lake Phewa. According to Hosking and Wallis [11] heterogeneity measure estimates the degree of heterogeneity in a group of sites and access whether they might reasonably treated as a homogeneous region. The heterogeneity measure (Equation 16) compares the observed and simulated dispersion of Lmoments for N sites under consideration. Hj
388
V
v j
j
v
,
j = 1, 2 and 3)
(16)
j
389 390 391
The regions are regarded as “acceptably homogeneous” when Hj < 1, “possibly heterogeneous” when 1 < Hj < 2, and “definitely heterogeneous” when Hj > 2. The details of the calculation of Hj are given in Hosking and Wallis [11].
392
2.6.3 Selection of Regional Distribution and Goodness of Fit
393 394 395 396 397 398 399
Lmoment ratio diagram were constructed using the unbiased estimators of Lmoments according to Hosking [89]. The curves show the theoretical relationships between LSkewness (L  Cs) and LKurtosis (L  Ck) of various candidate distributions. In addition, Zstatistics (ZDIST) defined by Hosking and Wallis [90] compare simulated L  Cs and L – Ck of fitted distribution with the regional average L  Cs and L – Ck values obtained from observed data. Following relation (Equation 17) defined the ZDIST;
400 401 402 403
Z DIST 4DIST t4R 4 4
(17) where, 4DIST is the LCk of fitted distribution, 4 and 4 simulated regional bias and simulated regional standard deviation of t4R . The simulation was made with the fitted Kappa distribution to regional Lmoments. The fit is regarded as adequate if Z DIST is close to zero and acceptable if
Z DIST 1.64 at confidence level of 90%. The satisfactory distributions were obtained as depicted in the Z DIST
404 405 406
diagram. Among the satisfactory distribution simpler distribution were chosen, accordingly regional growth curves were developed.
407
2.6.4 Estimation of Regional Growth Curves using IndexFlood Procedure
and Lmoment ratio
Hydrology 2018, 5, x FOR PEER REVIEW
408 409 410 411 412 413 414 415 416 417 418
12 of 31
The important assumption of the indexflood or the growth curves procedure was that the frequency distributions of N stations in the homogeneous region are identical apart from a sitespecific scaling factor of ith site can be written as (Equation 18): Qi ( F ) i q( F ), i 1, 2,.....N
(18) where, Qi ( F ) is the precipitation at site i at given return period, i is the indexflood [11], N is the number of sites and q ( F ) is the regional growth curve, a dimensionless quantile function common to homogeneous site. The indexflood i also known as scaling factor is the sample means of the data at site i. Since, we were concerned with the short duration extreme precipitation and thus hourly (extreme) mean precipitation was considered as indexflood ( i i ( hr ) ). Following equations (Equation 19 and 20) presents the regional growth curve defined by the quantile function of chosen distribution (GLO and GEV) for the western and eastern regions under consideration.
q( F )GLO
419
1 Tr 1
(19)
420 421 422 423 424 425 426 427 428
q( F )GEV
1 1 LOG 1 Tr
(20)
where, , and are respectively the scale, location and shape parameter and Tr is the return period. Utilizing the above two regional quantile function together with the i , we developed nominator [a(T)] of Equation 9. For the denominator [b(d)=(td+θ)η,] of Equation 9, we made a use of the reference IDF and solved the following (Equation 21) relation to obtain two parameters ( and ) of b(d) at given duration (td = 1 and 24 hours) by implementing SEP (Equation 12) as discussed in section 2.5.
i
429
a(T ) i ( hr ) .q( F ) b( d ) (td )
(21)
430
3. Results
431
3.1. Data Quality and Rainfall Variability
432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451
The evaluation of the historical daily rainfall series from the available 11 stations lead to the exclusion of three stations (Chapakot, Nr.810; Lamachaur, Nr.818 and Ghandruk, Nr. 821) from further analysis as these data passed the threshold of inhomogeneity or contained more than five percent of missing values. Therefore only eight stations were considered for further analysis. Table 1 presents the characteristics of the rainfall records, with missing values and homogeneity test results. The daily rainfall series of the eight stations showed that 81.40% of rainfall occurred during the monsoon season followed by premonsoon (11.10%), winter (4.0%) and postmonsoon (3.50 %). Out of the eight stations three are in the eastern part (BhadaureDeurali, PokharaAirport and Khairenitar), four in the western part (Syangja, Walling, KarkiNeta and Kusma) and Lumle is on the NW part of the Panchase hill range. The mean monsoonal rain of the eastern and western part was 2,657mm and 2,027mm respectively. The analysis also revealed that over a 30 year period in the eight stations the total number of storms exceeding 100mm in 24 hours was 998 of which 57 exceeded 200 mm (Fig. 3b). The highest recorded annual total precipitation in the region was 5,631 mm in 1984 in Lumle with the mean monsoonal sum of 4,681 mm (Fig. 3a) and the recorded maximum rain in 24 hours was 357 mm at PokharaAirport. This analysis clearly indicated that the rainfall in Panchase region is highly variable and the hill range distinctly divided the area into two meteorological regions. Over the 30 year period, the number of monsoonal dry days in the eastern part (BhadaureDeurali, PokharaAirport and Lumle) increased with overall upward trend whereas there was no such trend in the western part except in Syangja where number of dry days were found to be
Hydrology 2018, 5, x FOR PEER REVIEW
13 of 31
452 453 454 455
decreased with an increased number of storm events of more than 100 mm in 24 hours. We also observed that the annual mean monsoonal rainfall depth over the period of 30 year was more or less constant in all the stations. This suggests that the monsoonal rainfall intensity is increasing in the eastern part. Further to the west it was difficult to detect any clear trend.
456 457 458
Fig. 3. a) Seasonal rainfall pattern and b) monsoonal season extreme events and number of dry days (19822015). The numbers in the xaxis represents the stations.
459
3.2. Disaggregation of Daily Rainfall Depth
460 461 462 463 464 465 466 467 468
The rationale behind the rainfall disaggregation was to split the daily rainfall into hourly to subhourly [13, 39, 41, 42, 44, 71, 91] when the resolution of the available data is not fine enough. We derived MBLRP model parameters (Table 5) utilizing the statistical characteristics of historical monsoonal daily rainfall series (Table 4) for all eight stations and implemented the HyetosMinute software in R. We observed higher skewness value of daily rainfall series was possibly due to strong variability in the daily rainfall time series, whereas annual rainfall time series is rather smoothened out leading to low skewness, which was expected. The HyetosMinute effectively utilized the MBLRP model and disaggregated the monsoonal daily rainfall to synthetic hourly series even for long clusters of wet days separated by at least 1 dry day.
469 470
Table 4: Historical statistical characteristics of monsoonal daily rainfall series of the homogeneous data sets. Historical Characteristic s: Monsoon Season Mean (En’)
Lumle (Nr. 814)
KarkiNeta (Nr. 613)
BhadaureDeura
Kusma
Syangja
PokharaAi
Khairenitar
Walling
li (Nr.
(Nr.
(Nr.
rport (Nr.
(Nr.
(Nr.
813)
614)
805)
804)
815)
826)
38.36
17.40
27.89
17.46
18.63
25.81
14.27
13.59
1615.13
536.94
1099.75
527.67
840.10
1255.15
552.68
745.37
233.75
113.91
261.26
79.70
139.00
170.41
108.72
150.35
0.07
0.27
0.26
0.27
0.34
0.15
0.35
0.59
Skewness
1.65
2.13
1.96
1.92
2.69
2.45
2.81
3.28
Kurtosis
3.49
5.69
6.46
4.35
10.05
8.44
10.87
15.09
40.19
23.17
33.16
22.97
28.98
35.43
23.51
27.31
295.00
171.50
315.30
153.00
257.10
357.00
241.90
300.20
Variance (Varn’) Lag1 Covariance (Covn’) Percentage of Dry days (Pn’)
Standarddev iation Recorded Maximum
Hydrology 2018, 5, x FOR PEER REVIEW
14 of 31
Rain depth (mm day)
471 472
Table 5: Adopted MBLRP model parameters implemented in HyetosMinute derived from the historical monsoonal (daily) rainfall series of homogeneous data set of eight stations.
0.105
1.651
0.246
63.287
0.186
0.251
0.101
83.702
3.511
0.103
1.303
0.1493
55.347
1.031
9.904
0.470
0.577
0.171
68.614
805
0.863
5.797
0.224
1.145
0.201
68.704
804
1.601
4.505
0.304
1.022
0.686
74.703
Walling
826
0.423
6.816
0.200
1.895
0.129
58.762
Khairenitar
815
0.423
6.816
0.200
1.895
0.129
58.762
SN
Location
DHM (Nr.)
1
Lumle
814
2.282
4.039
2
KarkiNeta
613
0.939
4.327
3
BhadaureDeurali
813
1.087
4
Kusma
614
5
Syangja
6
Pokhara Airport
7 8
(day1)
(day)
x (mm/day)
473
3.3. Evaluation of Disaggregation Model
474 475 476 477 478 479 480 481 482
The model discussed in section 2.2 was evaluated utilizing the short duration hourly rainfall depth (JuneSept, 2016) available from the EPIC project. For this purpose, we implemented the global optimization algorithm and obtained the parameters discussed in section 2.3/2.4 according to Kossieris et al. [41] by comparing the original and disaggregated rainfall series in terms of statistical characteristics (e. g. mean, variance and skewness). We summarized the results for each month demonstrating that the disaggregated series preserved the statistical characteristics of the recorded rainfall series (Table 6) for the continuous weather station of the EPIC project at Gharelu in Kaski. However, we noticed the hourly synthetic rainfall series contained small differences in variance whereas the mean and skewness were exactly similar to that of original daily series.
483 484
Table 6: Comparison of statistical characteristics of recorded and disaggregated rainfall series respectively the mean, variance and skewness (En, Varn, Skewn and En’, Varn’, Skewn’) . Month
En
En’
Varn
Varn’
Skewn
Skewn’
June
15.41
15.41
461.53
461.52
1.644
1.644
July
46.07
46.07
2459.81
2459.79
1.105
1.105
August
28.50
28.50
1873.35
1873.36
1.831
1.831
September
45.67
45.67
1925.78
1925.77
0.966
0.966
485 486 487 488
In order to evaluate how well the model disaggregated the extreme intensity of the synthetic hourly rainfall depth we compared the recorded and disaggregated extreme intensity for each of the monsoonal months. The GoF test showed that there was a good agreement among the recorded and synthetic hourly intensities of the rainfall series ( 2 = 0.932, alpha=0.05 and degree of freedom = 3).
489
3.4. Selection of PDF and Parameter Estimation
490 491 492 493 494 495 496 497
PDF’s were generated for monsoonal daily and annual extremes rainfall series of eight stations using the statistical software EasyFit (section 2.4), focusing EV I and GEV models. The test statistics produced mixed result for annual extremes and monsoonal daily series. The KS and AD statistics ranked the three parameter ( , and ) GEV distribution model better over the EV I whereas the Chisquared test ranked the two parameter ( and ) EV I better than the GEV. The result of the analysis is shown in table 7.
Hydrology 2018, 5, x FOR PEER REVIEW
498 499
15 of 31
Table 7: PDF, estimated parameters and test statistics for the monsoonal daily time series and Annual extremes. Parameter: S. N.
1
2
3
Monsoonal Daily Location
Lumle
Karki Neta Bhadaure Deurali
4
Kusma
5
Syangja
6
7
8
Parameter:
PDF
GEV
Test Statistics
Monsoonal Annual
Time Series
Test Statistics
Extremes KS
AD 41.79
161.35
0.22
18.59
23.28
0.09
KS
AD
0.11
183.24
28.163
0.081
0.274
1.284
EV I

20.28
31.34
0.148
78.32
138.42

185.11
30.59
0.108
0.407
2.744
GEV
0.37
6.84
11.26
0.137
160.39
350.25
0.3
104.32
26.43
0.105
0.497
1.522
EV I

7.67
20.97
0.229
183.24
252.72

101.87
21.33
0.154
0.875
0.563
GEV
0.26
11.64
17.65
0.136
99.08
456.92
0.1
131.91
42.13
0.068
0.202
1.147
EV I

12.96
25.86
0.192
115.48
187.68

134.16
46.13
0.08
0.254
0.214
GEV
0.35
6.11
10.36
0.144
133.95
460.62
0.3
106.06
24.33
0.101
0.377
0.313
EV I

7.12
17.91
0.226
190.73
262.53

103.19
18.72
0.104
1.06
0.383
GEV
0.45
5.12
9.83
0.178
182.59
762.43
0.1
139.27
41.64
0.073
0.246
0.382
EV I

5.59
22.6
0.278
286.99
329.4

139.01
36.81
0.087
0.313
0.199
Pokhara
GEV
0.38
8.92
14.45
0.132
99.27
411.95
0.05
170.89
33.81
0.104
0.574
4.124
Airport
EV I

9.87
27.62
0.238
212.39
276.32

171.13
36.1
0.111
0.584
4.118
GEV
0.61
2.05
5.47
0.326
356.66
1804.8
0.11
126.09
33.66
0.134
0.372
0.08
Walling
Khairenitar
EV I

1.3
21.29
0.323
381.37
586.36

127.57
37.86
0.129
0.388
0.873
GEV
0.49
3.46
7.02
0.19
197.91
986.36
0.05
112.62
26.34
0.085
0.263
0.325
EV I

3.56
18.35
0.297
318.15
520.36

113.45
27.68
0.088
0.277
0.643
500 501 502 503 504 505 506
Stedinger et al. [82] expressed that the EV I distribution is obtained when = 0, and the general shape of GEV turns to EV I where < 0.3. In our case, for the monsoonal annual extremes the value was always less than 0.3 (Table 7) whereas this was not fully satisfied for the daily series. Also, according to Koutsoyiannis et al. [22] the performance of GEV is not very satisfactory with a small amount of samples as in our case. Based on the analysis at various levels (e. g. annual and monsoonal daily series, annual extremes and disaggregated hourly extremes series), we concluded that in this case the EV I distribution was better than GEV.
507
3.5. Construction of Reference and Empirical IDF Relationship
508 509 510 511 512 513 514 515 516
The reference IDF curves were developed for the coarser time resolution of 1 and 24 hours duration utilizing the extreme rainfall from the recorded daily and disaggregated hourly monsoonal rainfall series. The frequency of extreme rainfall depth (PT) and dimensionless Gumbel Frequency Factor (KT) also known as Gumbel Growth Curve [92, 93] were derived for the eight stations. The results showed that the frequency of precipitation is highest in BhadaureDeurali station and the lowest in Khairenitar (Fig. 4a). The figure (Fig. 4b) also shows the frequency factor (KT) for different return periods (Tr) which is similar for all stations. Utilizing the method of Wilson [85] we developed the IDF curves for durations of 1 and 24 hours, which we called the reference IDF curves (Fig. 6). These curves are the means to verify the finer time resolution IDF relationship.
517 518 519 520
Hydrology 2018, 5, x FOR PEER REVIEW
16 of 31
521
522 523 524
Fig. 4. a) Frequency of precipitation (PT) in an hour for all the stations; b) Dimensionless Gumbel Frequency Factor (KT) for 2, 5, 10, 25, 50, 100 and 200 year return period (Tr).
525 526 527
Fig. 5. Examples of computed reference IDF scenarios. a) KarkiNeta and b) PokharaAirport stations on loglog scale, td=1 and 24 hrs Tr=2, 5, 10, 25, 50, 100 and 200 Years).
528 529 530 531 532 533
In order to establish the IDF relationship for finer time resolution, the mathematical model proposed by Koutsoyiannis et al. [22] was parameterized. Accordingly the best fitted PDF parameters scale ( ) and location ( ) and other empirical constants and were estimated to
534 535
Table 8: Adopted parameters derived from EV I distribution ( , ) and estimated ( , )
536
be fitted into the mathematical model discussed in section 2.5. The estimated parameters are shown in table 8 which were used to develop stationspecific mathematical IDF relationship. Fig. 6 shows examples of the mathematically computed empirical IDF curves for the two stations.
according to Koutsoyiannis et al. [22] for monsoon season rainfall time series . SN
Location
DHM (Nr.)
1
Lumle
814
31.335
20.284
21.889
0.943
2
KarkiNeta
613
20.97
7.673
4.226
0.959
3
BhadaureDeurali
813
25.857
12.961
8.428
0.988
4
Kusma
614
17.91
7.125
4.38
0.98
5
Syangja
805
22.599
5.589
5.125
0.865
6
PokharaAirport
804
27.623
9.866
8.977
0.957
7
Walling
826
21.292
1.298
0.99
0.5
8
Khairenitar
815
18.354
3.564
0.988
0.472
Hydrology 2018, 5, x FOR PEER REVIEW
17 of 31
b)
a)
537 538 539 540
Fig. 6. Examples of mathematically computed IDF model in loglog scale for duration td (td = 5 min, 10 min, 30 min, 60 min, up to 24 hours) and return period Tr (Tr = 2, 5, 10, 25, 50, 100 and 200 years); a) PokharaAirport (Nr.: 804) and b) Syangja (Nr.: 805).
541
3.6. Evaluation of the IDF Relationship and Development of Empirical Model
542 543 544 545 546 547 548 549 550 551 552 553 554 555
The mathematically computed IDF relationship (section 3.5) was evaluated by fitting reference IDF scenarios and SEP and Chisquare tests were performed. The test statistics indicated that the mathematical model overestimated the IDF values for Lumle (Nr. 814) station for the both scenarios (i. e. 1 and 24 hour). Similarly Walling (Nr. 826) and BhadaureDeurali (Nr. 813) stations were poorly estimated for 1 hour duration. The possible reason for this could be either due to the PDF that we chose or the limitation of the mathematical model where intense and prolong rainfall occurs. This leads to the calibration of the model for which we again performed minimization of SEP by adjusting all four parameters ( , , and ) at a time. This reconstruction of the IDF significantly reduces
556 557
Table 9: Standard Error of Prediction (SEP) and Chisquare teststatistics (alpha = 0.05, test statistics = 12.59 for degree of freedom =7) before and after the calibration of the mathematical model.
the SEP and achieved significant test statistics except in Khairenitar station for which both SEP and GoF was increased but yet significant. Table 9 compared the test statistics before and after the calibration of the mathematical model. Also, examples of fitting the simulated IDF to that of the reference IDF are demonstrated in Fig. 7 in LOGLOG scale. Utilizing the calibrated empirical constant ( , ' , , ' ) '
'
and fitting them into the Equations 9, 10 and 11 we developed the
stationspecific empirical IDF model.
Before calibration DHM S.N.
Location
Station Number
After calibration
Standard Error
Chisquare Test
of Prediction
Statistics (alpha
(SEP)
at 0.05=12.59)
(Nr.)
Chisquare Standard Error of
Test Statistics
Prediction (SEP)
(alpha at
Region
0.05=12.59)
1 hr
24 hr
1 hr
24 hr
1 hr
24 hr
1 hr
24 hr
1
Lumle
814
6.59
8.22
7.55
24.36
0.10
0.00
0.00
0.00
2
KarkiNeta
613
3.19
2.10
1.81
3.55
2.05
0.00
0.72
0.00
3
BhadaureDeu
813
6.10
1.95
11.47
4.00
5.70
0.56
0.17
0.05
NW hill range Western part Eastern
Hydrology 2018, 5, x FOR PEER REVIEW
18 of 31 Before calibration
DHM S.N.
Location
Station Number
After calibration
Standard Error
Chisquare Test
of Prediction
Statistics (alpha
(SEP)
at 0.05=12.59)
(Nr.) 1 hr
24 hr
1 hr
Chisquare Standard Error of
Test Statistics
Prediction (SEP)
(alpha at 0.05=12.59)
24 hr
1 hr
24 hr
1 hr
24 hr
rali
part
4
Kusma
614
0.77
0.28
0.14
0.09
0.29
0.00
0.02
0.00
5
Syangja
805
1.27
0.52
0.24
0.21
0.00
0.87
0.00
0.60
804
1.37
1.21
0.36
0.97
0.60
0.90
0.06
0.55
Pokhara
6
Region
Airport
7
Walling
826
9.27
2.21
17.41
3.05
0.87
1.10
0.19
1.02
8
Khairenitar
815
4.98
0.44
3.96
0.14
0.64
0.63
6.55
8.77
Western part Western part Eastern part Western part Eastern part
558
559
Fig. 7. Examples of fitting reference IDF to the simulated IDF, a) PokharaAirport and b) Walling
560 561 562 563
Interpretation of the IDF reltionship indicated that the eastern part generally received higher rain than the western part. Variability was also observed in the rainfall intensity from east to west. Fig. 8 shows computed IDF relationship of four stations (Lumle, PokharaAirport, Syangja and Kusma) from the equations shown in table 10.
564
Table 10: Stationspecific empirical model based on adjusted parameters ( ' , ' , ' , ' ) . S. N.
Location
DHM Nr.
Empirical Model
i 5.469
7.047 ln ln 1
1
Khairenitar
815
2
Pokhara Airport
804
i 4.935
7.21 ln ln 1
3
BhadaureDeurali
813
i 5.953
5.77 ln ln 1
td 0.988 td 0.85
1 Tr
0.472
1 Tr
1 Tr
0.434
td 0.867
0.465
Hydrology 2018, 5, x FOR PEER REVIEW
S. N.
Location
19 of 31
DHM Nr.
Empirical Model
4
Lumle
814
5
Kusma
614
i 4.827
5.568 ln ln 1
6
Karkineta
613
i 7.003
4.775 ln ln 1
805
i 6.122
826
i 5.096
7
8
Syangja
Walling
565 a)
566
6.05 ln ln 1
i 6.197
b)
td 0.764 td 0.378
1 Tr
0.493
1 Tr
1 Tr
0.574
td 0.311
0.647
6.09 ln ln 1
td 0.998
1 Tr
0.501
6.093 ln ln 1
td 0.99
0.5
1 Tr
Hydrology 2018, 5, x FOR PEER REVIEW
20 of 31
d)
c)
567 568 569 570 571
Fig. 8. Graphical view of IDF relationships derived from the calibrated mathematical model ploted loglog scale for duration td (td = 5 min, 10 min, 30 min, 60 min, up to 24 hours) and return period Tr (Tr = 2, 5, 10, 25, 50, 100 and 200 years); a) Lumle (Nr. 814), b) PokharaAirport (Nr. 804), c) Syangja (Nr. 805) and d) Kusma (Nr. 614).
572 573 574 575 576 577 578 579 580
The theoretical probabilistic approach of the rainfall data distribution to compute the parameters of a(T) and optimization procedure for the parameters of b(d) that represent the dynamics of the rainfall pattern discussed in section 2.4 and 2.5 and demonstrated in Koutsoyiannis et al. [22] was the basis of this IDF relationship. However, there was some adjustment on the empirical constants that demonstrated better statistical significance and let the IDF curves passing through or much closer to the reference IDF point at given td and Tr. The reason of such adjustment was also explained in Koutsoyiannis et al. [22] and Van de Vyver and Demaree [17] who stated that the empirical considerations are not always consistent with the theoretical probabilistic approach of the IDF relationship.
581
3.7. Reganalization of IDF for Ungauged Locations
582 583 584 585 586 587 588
We executed RFA based on Lmoments for the heterogeneous regions divided into two homogeneous region (eastern region: Lumle, PokharaAirport, BhadaureDeurali and Khairenitar and western region: Kusma, KarkiNeta, Syangja and Walling) for which heterogeneity measure was evaluated. The heterogeneity measure of Lmoment demonstrated that the subregions were found to be statistically homogeneous. The estimated heterogeneity measures (H) for the eastern and western regions were  0.97 and 0.34 respectively. Similarly, the Lmoment ratio diagram and ZDIST demonstrated reasonably acceptable regional distribution functions (table 11 and Fig. 9).
589 590
Table 11: List of evaluated distribution indicating the GOF test statistics (ZDIST < 1.64 to accept the distribution). S. N.
Distribution
ZDISTStatistics /GOF Eastern Region
Western region
Remarks
1
Pearson Type III
0.19
1.31
accept
2
Gen. Normal
0.35
1.10
accept
3
Gaucho
0.48
2.37
accept/reject
4
Gen. Extreme Value
0.61
1.12
accept
5
Gen. Logistic
1.6
0.11
accept
Hydrology 2018, 5, x FOR PEER REVIEW
S. N. 6
21 of 31
ZDISTStatistics /GOF
Distribution
Eastern Region
Western region
1.66
3.67
Gen. Pareto
Remarks reject
591 592 593 594 595 596 597
For the development of regional growth curve we chose one set up simple but statistically acceptable distributions 1) Gen. Extreme Value (GEV) and 2) Gen. Logistic (GLO) respectively for the eastern and western region. Utilizing the empirical parameters (table 12) the quantile function (Equation 19 and 20) of the chosen distribution and hourly (extreme) mean precipitation (table 13) as an indexflood, we established the IDF relation for an hour (td = 1 hour) and given return period (Tr = 2, 5, 10, 25, 50, 100, 200 years).
598 599
Fig. 9. Lmoment ratio diagram for six distributions (note the black plus and square block representing the western region whereas the blue plus and red block is for eastern region).
600
Table 12: Chosen distribution their fitted and estimated empirical parameters. Parameters S. N.
601
Distribution
Location
Scale
Shape
( )
( )
( )
1
Gen. Logistics (GLO)
0.98
0.388
0.085
0.09
0.39
western
2
Gen. Extreme Value (GEV)
0.877
0.191
0.06126
0.51
0.40
eastern
Table 13: Mean hourly and daily (extreme) precipitation. Mean (extreme) S. N.
Location
hourly Precipitation (mm/hr)
602 603 604 605
Region
Mean (extreme) Daily Precipitation (mm/day)
Region
1
Khairenitar
14.09
6.60
eastern
2
PokharaAirport
25.29
10.90
eastern
3
BhadaureDeurali
25.35
11.30
eastern
4
Lumle
38.37
15.10
eastern
5
Syangja
18.69
7.80
western
6
KarkiNeta
17.07
7.00
western
7
Kusma
17.39
7.30
western
8
Walling
13.59
5.40
western
By solving the Equation 21 and from the regional distribution we obtained the regional parameters shown in table 12. We then evaluated the regional IDF with the station specific reference IDF for the given duration (td = 1 and 24 hours) and return period (Tr = 2, 5, 10, 25, 50, 100 and 200
Hydrology 2018, 5, x FOR PEER REVIEW
22 of 31
606 607 608 609 610 611
years) for which SEP and GOF was performed. The test depicted that the stations Khairenitar (Nr. 815), Lumle (Nr. 814) and Walling (Nr. 826 demonstrated insignificant IDF values [i.e. the regional model either over estimated or underestimated the IDF at the given duration (td = 1 hr and 24 hr)], indicating existence of some degree of heterogeneity, leading to the adjustment/calibration in the empirical parameters in table 12. In table 14, we presented the test statistics before and after the adjustment.
612 613
Table 14: Standard Error of Prediction (SEP) and Chisquare teststatistics (alpha = 0.05, test statistics = 12.59 for degree of freedom =7) before and after the calibration of regional model. Before calibration DHM S.N.
Location
Station Number
Standard Error of Prediction (SEP)
(Nr.)
614 615 616 617 618 619 620 621 622
After calibration
Chisquare Test Statistics
Chisquare
Standard Error
(alpha at
Test Statistics
of Prediction (SEP)
0.05=12.59)
0.05=12.59)
1 hr
24 hr
1 hr
24 hr
1 hr
24 hr
1 hr
24 hr
8.69
1.11
0.19
0.98
1
Khairenitar
815
117.6
24.8
0
0
2
Pokhara Airport
804
1.53
0.37
0.95
0.99
3
BhadaureDeurali
813
6.1
1.95
11.47
4
4
Lumle
814
37.15
17.43
0
0.01
5
Syangja
805
1.27
0.52
0.24
0.84
6
Kusma
614
1.37
1.21
0.36
0.86
7
KarkiNeta
613
9.27
2.21
17.41
0.85
8
Walling
826
37.54
8.52
0
0.20
Region
(alpha at
eastern eastern
no adjustment 9.34
0.11
0.16
eastern 0.99
eastern western
no adjustment
western western
24.60
3.60
0
0.73
western
The test statistics showed that the two homogeneous subregions (eastern and western) can be divided into further subregions such as three subregions in eastern region [eastern subregion 1: Khairenitar, eastern subregion 2: PokharaAirport and BhadaureDeurali and eastern subregion 3: Lumle] and two in western region [western subregion 1: Kusma, KarkiNeta and Syangja and western subregion and 2: Walling]. Utilizing the best fitted regional distribution and parameters we developed regional IDF formula for the region, shown in the table 15. Table 15: Lmoment based regional empirical IDF model for the Panchase region. S.
Area
Region/ DHM Nr.
N.
Empirical Model
(Distribution)
subregion
Khairenitar 1
815 (GEV)
Pokhara Airport & 2
804/813 BhadaureDeurali (GEV)
i
1 0.95 9.2 1 LOG 1 Tr
td 0.34
0.41
0.05
.
eastern1 hr
0.062 1 0.95 3.06 1 LOG 1 T r . i hr 0.4 td 0.5
eastern 2
Hydrology 2018, 5, x FOR PEER REVIEW
S.
23 of 31
Area
Region/ DHM Nr.
N.
Empirical Model
(Distribution)
subregion 0.062 1 0.88 3.06 1 LOG 1 T r . i hr 0.55 t 0.9 d
Lumle 3
814 (GEV)
614/613/805
0.98 4.47 1 Tr 1 i (td 0.09)0.39
826
0.99 14.86 1 Tr 1 i (td 0.01)0.35
Kusma/KarkiNeta & 4
Syangja (GLO) Walling
5 (GLO)
623
0.085
.
0.035
eastern3
western1 hr
.
western2 hr
td=duration in hour, Tr=return period in year and hr =hourly (extreme) mean precipitation
624 625 626 627
The adjustment of the empirical parameters helps to improve the test statistics except Walling, indicating that the distribution we chose was not feasible for western subregion  2. However, we could still say that the regional empirical model is useful to evaluate the IDF of rainfall for other subregions.
628
3.4. Rainfall Intensity in the Region
629 630 631 632 633 634
The computed rainfall intensity of eastern subregion3 was found to be the highest in terms of shorter duration rainfall (i. e. td = 0.5, 1 and 2 hrs), whereas for longer duration (i. e. td = 24hrs) the rainfall was relatively intense in western subregion 2 followed by eastern subregion 2 and western subregion 2. Among all, rainfall in eastern subregion 2 was less intense followed by western subregion 2 for shorter duration and return period. Table 16 presents the computed rainfall intensity of some important return periods (Tr) and duration (td) for example.
635 636
Table 16: Example of computed rainfall intensity (mm/hr) for some durations and return periods of region/subregion Duration (td in hr)
Return Period (Tr
Region/subregion
in Year)
0.5
1
2
24
5
31.6118
26.1030
20.7695
7.9508
25
45.4813
37.5555
29.8819
11.4392
100
57.8469
47.7663
38.0063
14.5493
5
28.8933
24.5675
20.0273
8.0377
25
45.4010
38.6037
31.4694
12.6299
100
54.7002
46.5107
37.9152
15.2168
5
43.2839
36.5917
28.9987
8.8877
25
55.8968
47.2544
37.4488
11.4775
100
67.3458
56.9333
45.1193
13.8284
5
32.8799
25.8801
20.0773
7.7383
25
50.5564
39.7935
30.8710
11.8985
100
66.5732
52.4005
40.6513
15.6681
eastern1
eastern2
eastern3
western1
Hydrology 2018, 5, x FOR PEER REVIEW
24 of 31 Duration (td in hr)
Return Period (Tr
Region/subregion
in Year)
0.5
1
2
24
5
29.7356
23.4107
18.3996
7.7230
25
47.0981
37.0801
29.1430
12.2325
100
61.6221
48.5147
38.1301
16.0047
western2
637
4. Discussion
638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680
Geographically the Panchase region can be divided into two subregions as eastern and western parts where the Panchase hill range is sitting in the midway extending NW to SE direction. This study demonstrated that the region can also divides into two parts in terms of rainfall variability. Monsoonal rainfall over the region is highly variable where the eastern part (i. e. Kaski District) received higher rain (>3,000mm, except Kahirenitar) than the western part (i. e. Syangja and Parbat Districts), where mean monsoonal rainfall was below 2,300mm. The geographical setting of the eastern part is relatively wider in comparison to the western part which is a valley containing several lakes (e. g. Phewa Lake, Begnas Lake and Rupa Lake, etc.). Rainfall is a complex process and the complexity is compounded due to the presence of lakes and high mountain topography. This complex process requires more detailed analysis considering wind direction; variation in daily temperature, solar radiation, etc. in order to better understand the rainfall variability. Use of the historical characteristics of the recorded daily rainfall to parametrize the disaggregation model to generate finer resolution synthetic rainfall series is an important step forward for data scarce regions [15, 39, 41, 45]. The method is useful especially for those locations where fine timeresolution rainfall data is not available leading to better estimating the finer resolution rainfall up to a minute [41]. However, we preformed the disaggregation procedure only for time periods of one hour or more since the possible error accumulation for finer resolution disaggregation is yet to be known, and has to be investigated (Kossieris, 2017:personal communication). Although many studies (e. g. [15, 37, 3942, 44, 45, 64, 94, 95] reported on the generation of synthetic rainfall series and their effectiveness, we noticed that the attention is less on the intensities. We observed that the synthetic rainfall intensities were unable to represent the recorded data of a particular day since the model implements a Poisson process and assumed that the intensity varies exponentially. In our case, however, the statistics suggest that the synthetic intensities were statistically significant for the monsoonal months and useful to develop reference IDF scenarios for the study region. In order to better understand the synthetic rainfall intensities, a rigorous analysis is required where long term fine resolution time series rainfall data are available. The application of Gumbel frequency factor technique is popular to evaluate frequencies of rain and flood storms [92, 93, 96, 97]. The technique is also frequently used to establish the IDF relationships [22, 93] using fine resolution rainfall data [22]. In our case due to inadequate data resolution we demonstrated the use of synthetic rainfall intensities and developed reference IDF curves utilizing the technique of Gumbel frequency factor as scenarios to be fitted into the mathematicallyderived empirical IDF model for better performance. In the literature, the use of scenario IDF computed from the disaggregated and recorded rainfall extremes to establish IDF relationship is limited. This is also the case for the approach of parameter adjustment in developing IDF relationships. The Lmoment based RFA method implemented in this study was able to demonstrate that the study area is heterogeneous in terms of geography and meteorology, what we observed while constructing the station specific IDF through Gumbel. This has leads to the division of the study area into subregions for the RFA. Application of Lmoment ratios and other statistics in identifying reasonably acceptable distributions of the data set is robust as it was depicted in the results. Therefore the technique is useful to better understand the IDF at the ungauged locations that can save computational time and resources. In this study, we observed that while fitting the regional IDF
Hydrology 2018, 5, x FOR PEER REVIEW
25 of 31
681 682 683 684 685 686 687
to the reference IDF, the chosen distribution slightly underestimated the intensity for the shorter duration and return period. However, in most cases the result was statistically significant and useful. The study indicated that the Panchase region is highly variable in terms of rainfall amount and intensity. Short duration and intense monsoonal rain may have different effects on floods and landslides than less intense but extended duration monsoonal rainfall respectively noticed in western and eastern subregions of the area.
688
5. Conclusions
689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731
This work demonstrated monsoonal rainfall variability over a geographic region in CentralWestern hills of Nepal and attempted to establish monsoon season IDF relationships where fine resolution rainfall data was scarce, common to many developing countries. The reported methodology and available tools are useful and applicable for ungauged locations within the study region. The available daily data were evaluated by grouping them in subsets according to the hydrological year. This approach was helpful to better understand the rainfall variability and better fitting of PDF’s as there was a strong seasonal effect on the rainfall in the region where more than 80% of rain occurred in the four monsoonal months. The Panchase hill geographically divides the region into an eastern (i. e. Kaski District) and western (i. e. Parbat and Syangja Districts) subregions. Analysis of above 30 years of homogeneous daily rainfall of eight stations in the region showed that the monsoonal rainfall amount was higher in the eastern part than that was recorded in western part. Also, the numbers of monsoonal dry days were found to be increased over the period of 30 years in the eastern part with constant amount of monsoonal rain indicated that the rainfall intensity is increased where in the western part no clear trend was noticed in the annual monsoonal rainfall amount except in Syangja where dry days are decreased with increased numbers of storms. However, because of geographical complexity to better understand the complex rainfall process in the region more detail analysis is needed. The proposed methodology to develop IDF relationship and empirical model at data scare situation was found statistically significant for which the available daily data were disaggregated to hourly synthetic series. The recorded daily and disaggregated hourly rainfall data were useful to compute the reference IDF for all the stations. The reference IDF was effective to evaluate the mathematically computed stationspecific IDF relationships. The mathematical model parameters ( , , and ) were estimated according to the established methods demonstrated by other researchers (e. g. [22]). Moreover the adjustment of the empirical constants performed better statistical significance of the IDF demonstrated usefulness of the model for the ungauged locations within the region. To regionalize the IDF relationships, the adopted Lmoment based RFA method was implemented dividing the region into two subregions first and later it was noticed that the region contains some degree of heterogeneity leading to produce five sets of empirical model. The models were used to estimate the regional IDF for the given duration and return period. Although the regional model under estimated the rainfall intensity for shorter duration and return period while fitting them into the reference IDF, were significant in most cases. This procedure can be extended for larger areas to establish IDF relationship in data scarce situation of Nepal leading to better management of storm water, road side drainage design, management and mitigation of various hazardsrisks induced by mass movement. The station specific (Gumbel) empirical IDF model demonstrated that the sort duration (i. e. 5, 10, 30, 60 min) rainfall intensity in the western part was higher than in the eastern part (highest in KarkiNeta) whereas this was not the case in the regional IDF model. The regional model revealed that the eastern subregion 2 (Lumle area) received the intense rain. The reason behind the different intensities was could be due to chosen distribution model. In addition to that the regional empirical IDF model generalized the distribution model to be fitted for the region and used the regional mean (extreme) precipitation leading to vary the intensities that come from the stationspecific IDF model. The interpretation of IDF clearly indicated that the Panchase hill range distinctly divides the region
Hydrology 2018, 5, x FOR PEER REVIEW
26 of 31
732 733 734
into two meteorological regions. The variability of the rainfall in terms of rain volume and intensity may have different effects on mass movement, soil erosion and flood in the region that have to be investigated for better and effective hazardrisk management.
735 736 737 738 739 740 741 742 743 744 745
Acknowledgements: We would like to acknowledge the Ecosystem Protecting Infrastructure and Communities (EPIC), project funded by the German Government, for funding this research and instrumentation especially allowing the installation and use of the continuously recording weather stations in Panchase region. We acknowledged IUCN for providing this research opportunity to the Department of Civil Engineering of Tribhuvan University of Nepal. Many thanks to the District Soil Conservation Office (DSCO) in Syangja, Kaski and Parbat Districts for their cooperation and support while in the field. Without the friendly cooperation from the Risks Analysis Group of University of Lausanne (UNIL), Switzerland, this study would have not been completed. Thanks are also due to Dr. Panagiotis Kossieris and Prof. Demetris Koutsoyiannis from the National Technical University of Athens and Dr. Govind Acharya an independent researcher from Nepal for providing literature, models and feedback on the manuscript.
746
References
747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775
1.
ADPC. Nepal Hazard Risk Assessment, Part 1. Bangkok, Thailand: Asian Disaster
Preparedness Center (ADPC), 2010. 2.
MoHA, DPNeT. Nepal Disaster Report 2015. Kathmandu, Nepal: DPNeT, 2015.
3.
MoENV. National Adaptation Porgramme of Action (NAPA), Ministry of Environmen. In:
Environment Mo, editor. Kathmandu, Nepal: Ministry of Environment 2010. p. 165. 4.
Petley DN, Hearn GJ, Hart A, Rosser N, Dunning S, Oven K, et al. Trends in Landslide
occurrence in Nepal. Natural Hazards 2007;43:2344. 5.
Petley DN, Hearn GJ, Hart A, Rosser N, Dunning S, Oven K, et al. On the impact of climate
change and population growth on the occurrence of landslides in Asia. Quarterly Journal of Engineering Geology and Hydrogeology. 2010;43(4):48796. 6.
CBS. Compendium of Environment Statistics of Nepal 2015. In: Statistics CBo, editor.
Kathmandu, Nepal: Government of Nepal; 2015. p. 176. 7.
Haigh M, Rawat JS. Landslide causes: Human impacts on a Himalayan landslide swarm,
Belgeo. Revue Balge de geographie. 2011:20120. 8.
Cruden DM, Varnes DJ. Landslide Types and Processes.
Landslide: Investigation and
Mitigation. 247. Denver, Colorado: Transportation Research Board, U.S. National Academy of Sciences, Special Report; 1996. p. 3675. 9.
SudmeierRieux K, Jaquet S, Derron MH, Jaboyedoff M, Devkota S. A Case Study of Landslides
and Coping Strategies in Two Villages of CentralEastern Nepal. Journal of Applied Geography. 2011;32:68090. 10. Jaboyedoff M, Derron MH, Voumard J, Leibundgut G, SudmeierRieux K, Nadim F, et al. Humaninduced landslides:
toward the analysis of
anthropogenic
changes
of
the
slope
environment. CRC Press. 2016. 11. Hosking JRM, Wallis JR. Regional Frequency Analysis: An Approach Based on LMoments. Cambridge; New York: Cambridge University Press; 1997. 12. Rakhecha PR, Singh VP. Applied Hydrometeorology. Capital Publishing Company, New Delhi, India: Springer, Dordrecht, The Netherlands; 2009. 384 p. 13. Rana A, Bengtsson L, Jothiprakash V. Development of IDFcurves for tropical India by random cascade modelling. Hydrology and Earth System Science Discussion. 2013;10.
Hydrology 2018, 5, x FOR PEER REVIEW
27 of 31
776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799
14. WMOUNESCOIAHS, editor Design of Water Resources Projects with inadequate data1973
800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818
Brazilian Archives of Biology and Technology. 2013.
1974; Madrid Symposium. 15. Ganguli
P,
Paulin
C.
Does
nonstationarity
in
rainfall
require
nonstationary
intensity–duration–frequency curves? Hydrology and Earth System Sciences 2017;21(12). 16. Nhat LM, Tachikawa Y, Taka K. Establishment of IntensityDurationfrequency Curves for Precipitaiton in the Monsoon area of Vietnam. Annuals of Disas Prev res. 2006:93103. 17. Van de Vyver H., Demaree G. R. Construction of IntensityDurationfrequency (IDF) curves for precipitation at Lubumabashi, Congo, under the hypothesis of inadequate data. Hydrological Sciences Journal. 2010;55
(4):55564.
18. Dahal RK, Hasegawa S. Representative rainfall thresholds for landslides in the Nepal Himalayas. Geomorphology 2008;100. 19. Langousis A, Veneziano D. IntensityDurationfrequency curves from scaling representation of rainfall. Water resources Research. 2007;43. 20. Van Asch TWJ, Buma J, Van Beek LPH. A view on some hydrological triggering system in landslides. Geomorphology 1999;30:2532. 21. Gerold LA, Watkins DW. Short Duration Rainfall Frequency Analysis in Michigan Using ScaleInvariance Assumptions. Journal of Hydrologic Engineering. 2005;10(6). 22. Koutsoyiannis D, Kozonis D, Manetas A. A mathematical framework for studying rainfall intensitydurationfrequency relationship. Journal of Hydrology. 1998;208. 23. Afrin S, Islam MM, Rahman MM. Development of IDF Curves for Dhaka City based on Scaling Theory under future precipitation variability due to climate change. International Journal of Environmental Sciences and developmen. 2015;6(5). 24. Cardoso CO, Bertol I, J. SO, de Paiva Sampaio CA. Generation of Intensity Duration frequency Curves and Intensity temporal variability Pattern of Intense rainfall for Lages
25. Verhoest N, Troch PA, De Troch FP. On the applicability of BartlettLewis rectangular pulses models in the modelling of design storms at a point. Journal of Hydrology 1997;202:10820. 26. Chen CL. RainfallIntensityDuration Formulas. ASCE Journal of Hydrology and Engineering. 1983. 27. Baghirathan VR, Shaw EM. Rainfall depthdurationfrequency studies for Sri Lanka. J Hydrol. 1978;37: 223–39. 28. Gert
A,
Wall
DJ,
White
EL,
Dunn
CN.
REGIONAL
RAINFALL
INTENSITYDURATIONFREQUENCY CURVES FOR PENNSYLVANIA. Journal of American Water Resources. 1987. 29. Kothyari UC, Grade RJ. Rainfall intensity duration frequency formula for India. J Hydr Engrg. 1992;118. 30. Yu PS, Chen CL. regional analysis of rainfall intensitydurationfrequency relationship Jounral of Chinese Institute of Engineers 1996;19(4):52332. 31. Madsen H, Mikkelsen PS, Rosbjerg D, Harremoes P. Regional estimation of rainfall intensitydurationfrequency curves using generalized least squares regression of partial duration series statistics. Water Resources Research. 2002;38(11):2131. 32. Willems P. Compound intensitydurationfrequency relationships of extreme precipitation for two seasons and two storm types Journal of Hydrology. 2000;233(1):189205.
Hydrology 2018, 5, x FOR PEER REVIEW
28 of 31
819 820 821 822 823 824 825 826 827
33. Yu PS, Yang TC, Lin CS. Regional rainfall intensity formulas based on scaling property of
828 829 830 831
38. Glasbey CA, Cooper G, McGechan MB. Disaggregation of daily rainfall by conditional
832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859
40. Kilsby CDG, Jones PD, Burton A, Ford AC, Fowler HJ, Harpham C, et al. A daily weather
rainfall. Journal of Hydrology. 2004;295(14):10823. 34. Dalrymple T. Dalrymple, Tate. Floodfrequency analyses. USGPO, 1960 1960. Report No. 35. Ariff NM, Jemain AA, Bakar MAA. Regionalization of IDF Curves with LMoments for Storm Events. International Journal of Mathematical and COmputation Sciences. 2016;10(5):21723. 36. Hosking JRM, Wallis JR, Wood EF. An appraisal of the regional frequency analysis Hydrological Sciences Journal. 1985;30(1):85109. 37. Gaume E, Mouhous N, Andrieu H. Rainfall stochastic disaggregation models: Calibration and validation of a multiplicative cascade model. Adv Water Resources. 2007;30.
simulation of pointprocess model. Jounral of Hydrology. 1995;165. 39. Kaczmarska J, Isham V, Onof C. Point process models for fineresolution rainfall. Hydrological Science Journal. 2014;59(11):197291.
generator for use in climate change studies. Journal of Environmental Modelling and Software. 2007;22:170519. 41. Kossieris P, Makropoulos C, Onof C, Koutsoyiannis D. A rainfall disaggregation scheme for subhourly time scales: Coupling a BartlettLewis based model with adjusting procedures. Journal of Hydrology. 2016;556:98092. 42. Koutsoyiannis D. A stochastic disaggregation method for design storm and flood synthesis. Journal of Hydrology. 1994;156(14):193225. 43. Koutsoyiannis D, Onof C. A computer program for temporal rainfall disaggregation using adjusting procedures (HYETOS). XXV General Assembly of European Geophysical Society; Nice, France2000. 44. Koutsoyiannis D, Onof C. Rainfall disaggregation using adjusting procedures on a Poisson cluster model. Journal of Hydrology. 2001;246(14):10922. 45. Kaczmarska J. Further Development of BarlettLewis model for fineresolution rainfall. . Department of Statistical Science, University College London. 2011. 46. Shrestha ML. Interannual variation of summer monsoon rainfall over Nepal and its relation to Southern Oscillation Index. Meteorology and Atmospheric Physics. 2000;75(12):218. 47. Chalise SR, Khanal NR, editors. An introduction to climate, hydrology and landslide hazards in the Hindu KushHimalayan region. In Landslide hazard mitigation in the Hindu KushHimalaya; 2001; Kathmandu, Nepal: ICIMOD. 48. Leibundgut G, SudmeierRieux K, Devkota S, Jaboyedoff M, Derron MH, Penna I, et al. Rural earthen roads impact assessment in Phewa watershed, Western region, Nepal. Geoenvironmental Disasters. 2016;3(1). 49. Devkota S, Adhikari BR. Development of ecosystem based sediment control techniques and design of siltation dam to protect Phewa Lake, Kaski District, Nepal. Kathmandu, Nepal: GOVN/UNDP/IUCN; 2015. 46 p. 50. Fleming B, Felming PJ. A watershed conservation success story in Nepal: Land use changes over 30 years. Waterlines. 2009;28(1):2946.
Hydrology 2018, 5, x FOR PEER REVIEW
860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902
29 of 31
51. Gurung H. Pokhara Valley: A field study in regional geography. Edingurg, UK: University of Edingurg; 1965. 52. JICA. The Development study on the environmental conservation of Phewa Lake. Unpublished project report. Kathmandu, Nepal: JICA, 2002. 53. Thapa GB, Weber KE. Status and management of watersheds in the upper Pokhara valley, Nepal. Environmental Management 1995;19(4):497513. 54. Buishand TA. Some methods for testing the homogeneity of rainfall records. Hydrology. 1982;58(12):1127. 55. Wijngaard JB, Klein Tank AMG, Können GP. Homogeneity of 20th century European daily temperature and precipitation series. International Journal of Climatology. 2003;23(6):67992. 56. Raes D, Willems P, GBaguidi F, editors. RAINBOW – a software package for analyzing data and testing the homogeneity of historical data sets. 4th International Workshop on ‘Sustainable management of marginal drylands'; 2006; Islamabad, Pakistan. 57. RClimTool. RClimTool: A free application for analyzing climatic series. . Adaptacion para la Sostenibilidad Productiva: Clima y Sector Agropecuario Colombiano; 2014. 58. XLSTAT. XLSTAT Getting started manual. XLSTAT; 2016. 59. Devkota S, Lal AC. Local Knowledge for Addressing Climate Change Risks at Local level In: SudmeierRieux K, Fernandez, M., Penna, I. M., Jaboyedoff, M., and Gaillard, J. C, editor. Identifying Emerging Issues in Disaster Risk Reduction, Migration, Climate Change and Sustainable Development Springer; 2017. 60. Koutsoyiannis D, Manetas A. Simple disaggregation by accurate adjusting procedures. Water Resources Research. 1996:32. 61. RodriguezIturbe I, Cox DR, Isham V, editors. Some models for rainfall based on stochastic point processes. Mathematical, Physical and Engineering Sciences 1987; London: Proceedings Royal Society. 62. Onof C, Wheater HS. Improvement to the modelling of British rainfall using a modified random parameter BartlettLewis rectangular Pulses Model. Journal of Hydrology. 1994;157(14):17795. 63. RodriguezIturbe I, Cox DR, Isham V, editors. A point process model for rainfall: Further developments. Mathematical, Physical and Engineering Sciences 1988; London: Royal Society A. 64. Abdellatif M, Atherton W, Alkhaddar R. Application of the stochastic model for temporal rainfall disaggregation for hydrological studies in north western England. J Hydroinformatics 2013. 65. Cowpertwait PSP, O’Connel PE, Matcalfe AV, Mawdsley JA. Stochastic point process modelling of rainfall, I, Single site fitting and validation. Journal of Hydrology. 1996a;176. 66. Cowpertwait PSP, O’Connel PE, Matcalfe AV, Mawdsley JA. Stochastic point process modelling of rainfall. II, Regionalization and disaggregation. Journal of Hydrology. 1996b;176. 67. Islam D, Entekhabi D, Bras RL. Parameterestimation and sensitivity analysis for the modified BartlettLweis rectangular pulses model of rainfall. Journal of Geophysics Research. 1990;95. 68. Onof C, Wheater HS. Modelling of British rainfall using a random parameter BartlettLewis rectangular Pulse Model. Journal of Hydrology. 1993;149(14):6795. 69. Onof C, Wheater HS. Improved fitting of the BartlettLewis Rectangular Pulse Model for hourly rainfall. Hydrological Sciences Journal. 1994;39(6):66380. 70. Pui A, Sharma A, Mehrotra R, Sivakumar B, Jeremiah E. A comparison of alternatives for daily to subdaily rainfall disaggregation. Journal of Hydrology. 2012;470471:13857.
Hydrology 2018, 5, x FOR PEER REVIEW
30 of 31
903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942
71. Yusop HZ, Yusof F. The use of BLRP Model for Disaggregating Daily Rainfall Affected by
943 944
90. Hosking JRM, Wallis JR. Some statistics useful in regional frequency analysis Water Resources
Monsoon in Peninsular Malaysia. Sains Malaysian. 2016;45(1):8797. 72. RProject. ‘HyetosMinute’ Rainfall Disaggregation Software Plug in Package of R: temporal stochastic simulation of rainfall at fine time scale In: Kossieris P, Tyrals H., Koutsoyaiannis D., Makropoulos C., and Efstratiadis A., editor.: The R Project for Statistical Computing. ; 2016. 73. Ghosh S, Roy MK, Biswas SC. Determination of the best fit probability distribution for monthly rainfall data in Bangladesh. American Journal of Mathematics and Statistics. 2016;6(4):5. 74. Hanson LS, Vogel R. The probability distribution of daily rainfall in the United States. World Environmental and Water Resources Congress 2008. 75. Millington N, Das S, Simonovic SP. The Comparison of GEV, LogPearson Type 3 and Gumbel Distributions in the Upper Thames River Watershed under Global Climate Models. THE UNIVERSITY OF WESTERN ONTARIO DEPARTMENT OF CIVIL AND ENVIRONMENTAL ENGINEERING, 2011 Contract No.: 077. 76. Svensson C, Jones DA. Review of rainfall frequency estimation methods. Journal of Flood Risk Managment 2010;3(4):296313. 77. Win NL, Win KM. The Probability Distributions of Daily Rainfall for Kuantan River Basin in Malaysia. International Journal of Science and Research 2014;3(8). 78. EasyFit. Data analysis and Simulation, Distribution Fitting Tutorials. 2016. 79. Gamage SHPW, Hewa GA, Beecham S. Probability distributions for explaining hydrological losses in South Australian catchments. Hydrology and Earth System Sciences. 2013;17(11):454153. 80. Misic B. Communication in large scale basin Networks in theory and application: University of Toronto; 2014. 81. Hosking JRM, Wallis JR, Wood EF. Estimation of the generalized extremevalue distribution by the method of probabilityweighted moments. Technometric. 1985;27(3):11. 82. Stedinger JR, Vogel RM, FoufoulaGeorgiou E. Frequency analysis of extreme events. In: Maidment DR, editor. Handbook of Hydrology. New York: McGrawHill, NY; 1993. 83. Koutsoyiannis D. Statistics of extremes and estimation of extreme rainfall, Theoretical investigation. Hydrological Sciences Journal. 2004a;49(4):57590. 84. Koutsoyiannis D. Statistics of extremes and estimation of extreme rainfall, Empirical investigation of long rainfall records. Hydrological Sciences Journal. 2004b;49(4):591610. 85. Wilson EM. Engineering Hydrology. fourth edition ed. London: Macmillan, London. ; 1990. 86. Benabdesselam T, Amarchi H. Regional approach for the estimation of extreme daily precipitation on Northeast area of Algeria. 2013. 87. Burn DH. Evaluation of regional Flood Frequency Analysis with a region of Influence Approach. Water Resources Research. 1990;26(10):225765. 88. Greenwood JA, Landwehr JM, Matalas NC, Wallis JR. Probability weighted moments: definition and relation to parameters of several distributions expressible in inverse form. Water Resources Research. 1979;15:104954. 89. Hosking JRM. LMoments: Analysis and Estimation of Distributions Using Linear Combinations of Order Statistics. Journal of the Royal Statistical Society Series B 1990;52 (1):10524
Research 1993;29(2):27181.
Hydrology 2018, 5, x FOR PEER REVIEW
31 of 31
945 946 947 948 949 950 951 952 953 954 955 956 957 958 959
91. Villani V, di Serafino D, Rianna G, Mercogliano P. Stochastic models for the disaggregation of
960
© 2018
precipitation time series on subdaily scale: identification of parameters by global optimization. Research Papers. Centro Euromediterraneo sui Cambiamenti Climatici, 2015 Contract No.: RP0256 92. AlHassoun AS. Developing an empirical formulae to estimate rainfall intensity in Riyadh region. Journal of King Saud University  Engineering Sciences. 2011;23(2):818. 93. DePaola F, Giugni M, Topa ME, Bucchignani E. IntensityDurationFrequency (IDF) Curves, for data series and climate projection in African cities. Springerplus. 2014;3:133. 94. Connolly RD, Schirmer J, Dunn PK. A daily rainfall disaggregation model. Agr Forest Meteorolog. 1998;92. 95. Koutsoyiannis D, Onof C, Wheater HS. Multivariate rainfall disaggregation at a fine timescale. Water Resources Research. 2003;39(7). 96. Apel H, Thieken HA, Merz B, Blöschl G. A Probabilistic Modelling System for Assessing Flood Risks. Natural Hazards. 2006;38(12):79100. 97. Solomon O, Prince O. Flood Frequency Analysis using Gumbel’s Distribution. Civil and Environmental Research. 2013;3(10). by the
961 962
authors. Submitted for possible open access publication under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).