Temporal Rainfall Disaggregation Research Articles

Multisite temporal rainfall disaggregation using methods of fragments conditioned on circulation patterns

A multisite method of fragments-based rainfall temporal (daily to hourly) disaggregation procedure conditioned on circulation pattern (CP) classification is developed and applied to the German part of the Rhine river basin. The performance of the CP-based disaggregation in representing key rainfall characteristics is evaluated and the influence of different number of CP classes is investigated. Compared with the standard (monthly-based) method of fragments disaggregation, the CP-based procedure including a seasonal stratification improves the disaggregation quality in hourly rainfall extremes, although it shows no improvement in reproducing standard rainfall statistics, such as mean and standard deviation. CP classifications with more classes tend to perform better in representing hourly rainfall statistics in the Rhine basin. The developed procedure opens up the possibility to consider dynamic changes in the atmosphere, such as changes in the frequency, persistence and seasonality of circulation patterns, that may be associated with climate change.

Predicting flood events in Kathmandu Metropolitan City under climate change and urbanisation

Urbanisation and climate change collectively impose the threat of urban flood. The impervious transformation and changes in local climatic conditions increase the risk of frequent pluvial flooding in Kathmandu. Therefore, this study aims to assess the integrated impact of urbanisation and climate change on pluvial flooding in Kathmandu Metropolitan City, using the Personal Computer Storm Water Management Model. The future daily rainfall from three Regional Climate Models under two Representative Concentration Pathways (RCPs) and the observed daily data has been disaggregated to hourly series using temporal rainfall disaggregation. The result shows that the combined impact on pluvial flood is likely to intensify with relatively more contribution from climate change. The CNRM-CM5-CSIRO-CCAM under the RCP 4.5 scenario projects the maximum increase (60–90%) in flood volume for current (75%) and extreme (90%) imperviousness, under a 2-year and 20-year return period (RP) with the extent of the flood area increasing more than the depth. For a 2-year RP, areas with a depth of 0.10–0.25 m are likely to expand while those from 0.25 to 0.40 m in depth are projected to have more velocity, and these trends are expected to be magnified for higher RPs. The study shows that even though urbanisation contributes less to urban pluvial flood, it can be a catalyst for exaggerating other factors and implementing management measures. For instance, the application of small-scale rainwater harvesting, and overflow storage reduced the runoff, thereby reducing the flood volume by 20–35%. Furthermore, the findings highlight the need for planning management strategies such as evaluation and upgrading of conventional drainage systems with low impact measures (rainwater harvesting and green roofs); technical and financial assistance to urban dwellers in adopting the management measures or a combination of them based on the location, requirements, and availability of information and resources to reduce the effects of pluvial flood in Kathmandu.

Temporal rainfall disaggregation using a micro-canonical cascade model: possibilities to improve the autocorrelation

Abstract. In urban hydrology rainfall time series of high resolution in time are crucial. Such time series with sufficient length can be generated through the disaggregation of daily data with a micro-canonical cascade model. A well-known problem of time series generated in this way is the inadequate representation of the autocorrelation. In this paper two cascade model modifications are analysed regarding their ability to improve the autocorrelation in disaggregated time series with 5 min resolution. Both modifications are based on a state-of-the-art reference cascade model (method A). In the first modification, a position dependency is introduced in the first disaggregation step (method B). In the second modification the position of a wet time step is redefined in addition by taking into account the disaggregated finer time steps of the previous time step instead of the previous time step itself (method C). Both modifications led to an improvement of the autocorrelation, especially the position redefinition (e.g. for lag-1 autocorrelation, relative errors of −3 % (method B) and 1 % (method C) instead of −4 % for method A). To ensure the conservation of a minimum rainfall amount in the wet time steps, the mimicry of a measurement device is simulated after the disaggregation process. Simulated annealing as a post-processing strategy was tested as an alternative as well as an addition to the modifications in methods B and C. For the resampling, a special focus was given to the conservation of the extreme rainfall values. Therefore, a universal extreme event definition was introduced to define extreme events a priori without knowing their occurrence in time or magnitude. The resampling algorithm is capable of improving the autocorrelation, independent of the previously applied cascade model variant (e.g. for lag-1 autocorrelation the relative error of −4 % for method A is reduced to 0.9 %). Also, the improvement of the autocorrelation by the resampling was higher than by the choice of the cascade model modification. The best overall representation of the autocorrelation was achieved by method C in combination with the resampling algorithm. The study was carried out for 24 rain gauges in Lower Saxony, Germany.

TEMPORAL RAINFALL DISAGGREGATION BY A SIMPLE RANDOM CASCADE MODEL

This study evaluates the utility and suitability of a simple discrete multiplicative random cascade model for temporal rainfall disaggregation. Two of a simple random cascade model, namely log-Poisson and log-Normal models are applied to simulate hourly rainfall from daily rainfall at seven rain gauge stations in Peninsular Malaysia. The cascade models are evaluated based on the capability to simulate data that preserve three important properties of observed rainfall: rainfall variability, intermittency and extreme events. The results show that both cascade models are able to simulate reasonably well the commonly used statistical measures for rainfall variability (e.g. mean and standard deviation) of hourly rainfall. With respect to rainfall intermittency, even though both models are underestimated, the observed dry proportion, log-Normal model is likely to simulate number of dry spells better than log-Poisson model. In terms of rainfall extremes, it is demonstrated that log-Poisson and log-Normal models gave a satisfactory performance for most of the studied stations herein, except for Dungun and Kuala Krai stations, which both located in the east part of Peninsula.

Impact of Climate Change on the Development of Rainfall Intensity, Duration and Frequency Curves in <i>Chiro</i> and <i>Hurso</i> Stations of Eastern Ethiopia

Today environmental issue becomes the biggest concern of humankind because of scientific evidence about the increasing concentration of greenhouse gases in the atmosphere and the changing climate of the Earth. This study was conducted in eastern Ethiopia specifically at <i>Chiro</i> and <i>Hurso</i> stations. The study assessed quantitatively the rainfall Intensity Duration Frequency (IDF) relationships under changing climate condition and compare with the existing rainfall- Intensity Duration- Frequency (IDF) relationships. Rainfall intensity duration and frequency curves were developed using historical rainfall time series data under the assumption that climate is stationary. This assumption is not valid under changing climatic conditions that may shifts in the magnitude and frequency of extreme rainfall. Such shifts in extreme rainfall at the local level demand new regulations for any intervention management as well as changes in design practices. In order to estimate the level of climate change impact on the rainfall Intensity Duration Frequency (IDF) relationships, these changes of the climate variables were applied to Hyetos Temporal Rainfall Disaggregation model to simulate future IDF relationships. From the results can see graphical presentation of IDF curves for return periods of 2, 5, 10, 25, 50 and 100 years for durations of 1, 2, 3, 6, 12 and 24 hours. The comparison results indicate that, difference between rainfall intensities (percentage) of climate change scenario and historic rainfall for 2020s ranges between 1.58% and 10.92% for 2050s, ranges between 0.07% and 20.22% and for 2080s, ranges between 0.71% and 55.93% in <i>Chiro</i> station, respectively. Similarly, in the case of <i>Hurso</i> station, the difference between climate change scenario and historic rainfall for 2020s ranges between1.10% and 27.83% for 2050s ranges between 110.5% and 40.21% and for 2080s that ranges between 19.44% and 67.75%, respectively. Therefore, the outputs of the study indicates that the rainfall magnitude will be different in the future and thereby the decrease and increase in rainfall intensity and magnitude may have major implications on ways in which current and future intervention is designed, operated, and maintained. Therefore, design standards and guidelines currently employed in the study area should be revised with the confirmation of the impacts of climate change.

Temporal rainfall disaggregation using a multiplicative cascade model for spatial application in urban hydrology

Rainfall time series of high temporal resolution and spatial density are crucial for urban hydrology. The multiplicative random cascade model can be used for temporal rainfall disaggregation of daily data to generate such time series. Here, the uniform splitting approach with a branching number of 3 in the first disaggregation step is applied. To achieve a final resolution of 5 min, subsequent steps after disaggregation are necessary. Three modifications at different disaggregation levels are tested in this investigation (uniform splitting at Δt = 15 min, linear interpolation at Δt = 7.5 min and Δt = 3.75 min). Results are compared both with observations and an often used approach, based on the assumption that a time steps with Δt = 5.625 min, as resulting if a branching number of 2 is applied throughout, can be replaced with Δt = 5 min (called the 1280 min approach). Spatial consistence is implemented in the disaggregated time series using a resampling algorithm. In total, 24 recording stations in Lower Saxony, Northern Germany with a 5 min resolution have been used for the validation of the disaggregation procedure. The urban-hydrological suitability is tested with an artificial combined sewer system of about 170 hectares. The results show that all three variations outperform the 1280 min approach regarding reproduction of wet spell duration, average intensity, fraction of dry intervals and lag-1 autocorrelation. Extreme values with durations of 5 min are also better represented. For durations of 1 h, all approaches show only slight deviations from the observed extremes. The applied resampling algorithm is capable to achieve sufficient spatial consistence. The effects on the urban hydrological simulations are significant. Without spatial consistence, flood volumes of manholes and combined sewer overflow are strongly underestimated. After resampling, results using disaggregated time series as input are in the range of those using observed time series. Best overall performance regarding rainfall statistics are obtained by the method in which the disaggregation process ends at time steps with 7.5 min duration, deriving the 5 min time steps by linear interpolation. With subsequent resampling this method leads to a good representation of manhole flooding and combined sewer overflow volume in terms of hydrological simulations and outperforms the 1280 min approach.

Temporal Rainfall Disaggregation with a Cascade Model: From Single-Station Disaggregation to Spatial Rainfall

Temporal rainfall disaggregation is an important tool to obtain high-resolution rainfall data, which are needed in many fields of hydrology and water resources management. The multiplicative random cascade model can be used for temporal rainfall disaggregation of daily time series. A resampling algorithm is introduced to implement spatial consistence in disaggregated time series. Spatial consistence is assumed to be represented by four bivariate and distance-dependent rainfall characteristics that complement each other. Relative diurnal cycles of the disaggregated time series are resampled with the aim to reproduce these spatial characteristics while preserving the structure generated by the cascade model. Also, to achieve a final resolution of 1 h the traditional cascade model has been modified. A modification called uniform splitting with a branching number of 3 in the first step is introduced. Results are compared with observations and an approach called diversion. In total 22 recording stations in Northern Germany with hourly resolution were used for the validation of the disaggregation procedure, starting with daily values. Investigation areas are two catchments considering different station densities. The results show that for the disaggregation, errors of time series characteristics between 3 and 12% occur. The nonexceedance curves of rainfall intensities are slightly overestimated. Extreme values are well represented. The uniform splitting method outperforms the diversion method. Spatial rainfall characteristics can be reproduced by the simulating annealing algorithm. However, with an increasing number of stations the reproduction performance declines for some rainfall characteristics. Nonexceedance curves of areal rainfall based on disaggregated and not resampled time series are generally underestimated. By application of the resampling algorithm, a better performance regarding the spatial characteristics can be achieved. The presented resampling algorithm also has the potential to be used for implementing spatial consistence for time series generated by other disaggregation models.

Circulation pattern based parameterization of a multiplicative random cascade for disaggregation of observed and projected daily rainfall time series

Abstract. The use of multiplicative random cascades (MRCs) for temporal rainfall disaggregation has been extensively studied in the past. MRCs are appealing for rainfall disaggregation due to their formal simplicity and the possibility to extract the model parameters directly from observed high resolution rainfall data. These parameters, however, represent the rainfall characteristics of the observation period. Since rainfall characteristics of different time slices are changing due to climate variability, we propose a parameterization approach for MRCs to adjust the parameters according to past (observed) or future (projected) time series. This is done on the basis of circulation patterns (CPs) by extracting a distinct MRC parameterization from high resolution rainfall data, as observed on days governed by each individual CP. The parameterization approach is tested by comparing the statistical properties of disaggregated rainfall time series of two time slices, 1969–1979 and 1989–1999, to the results obtained by two other disaggregation methods (a conceptually similar MRC without CP-based parameterization and a recombination approach) and to the statistical properties of observed hourly rainfall data. In this context, all three approaches use rainfall data of the time slice 1989–1999 for parameterization. We found that the inclusion of CPs into the parameterization of a MRC yields hourly time series that better reproduce the properties of observed rainfall in time slice 1989–1999, as compared to the simple MRC. Despite similar results of both MRCs in the validation period of 1969–1979, we can conclude that the CP-based parameterization approach is applicable for temporal rainfall disaggregation in time slices distinct from the parameterization period. This approach accounts for changes in rainfall characteristics due to changes in the frequency of occurrence of the CPs and allows generating hourly rainfall from daily data, as often provided by a statistical downscaling of global climate change.

Application of the stochastic model for temporal rainfall disaggregation for hydrological studies in north western England

Assessment of climate change on any hydrological system requires higher temporal resolution at hourly or less in terms of time scale. This paper implements the Bartlett–Lewis Rectangular Pulses (BLRP) model coupled with a proportional adjusting procedure to disaggregate daily rainfall to hourly rainfall in order to demonstrate the reliability of this method. Three stations in northwestern England have been selected that represent different climates in the region. Parameters estimation of the BLRP model has been performed under different levels of hourly rainfall aggregation for a combination of rainfall statistics. The HYETOS model, which applies BLRP, reproduced standard statistics such as mean, variance, Lag-1, autocorrelation as well as dry proportions. Moreover, the model was proven to have the capability to disaggregate the rainfall extremes. The fitted BLRP model could then be used to disaggregate future daily rainfall in order to investigate the climate change impact of different rainfall intensities.

Time scale and intensity dependency in multiplicative cascades for temporal rainfall disaggregation

Multiplicative random cascades (MRCs) can parsimoniously generate highly intermittent patterns similar to those in rainfall. The elemental MRC model parameter is the cascade weight, which determines how rainfall at one scale is partitioned at the next smallest scale in the cascade. While it is known that the probability density of these weights may vary with both time scale and rainfall intensity, nearly all previous studies have considered either time scale or intensity separately. We examined the simultaneous dependency of the weights on both factors and assessed the impacts of explicitly including these dependencies in the MRC model. On the basis of the observed relationships between cascade weights and time scale and intensity, four progressively more “dependent” models were constructed to disaggregate a long time series of daily rainfall to hourly intervals. We found that inclusion of the intensity dependency on the model parameters that generate dry intervals greatly improved performance. For the relatively small range of time scales over which the rainfall was disaggregated, varying model parameters with time scale resulted in minor improvement.

Evaluation of an artificial neural network rainfall disaggregation model

Previous research produced an artificial neural network (ANN) temporal rainfall disaggregation model. After proper training the model can disaggregate hourly rainfall records into sub-hourly time increments. In this paper we present results from continued evaluations of the performance of the ANN model specifically examining how the errors in the disaggregated rainfall hyetograph translate to errors in the prediction of the runoff hydrograph. Using a rainfall-runoff model of a hypothetical watershed we compare the runoff hydrographs produced by the ANN-predicted 15-minute increment rainfall pattern to runoff hydrographs produced by (1) the observed 15-minute increment rainfall pattern, (2) the observed hourly-increment rainfall pattern, and (3) the 15-minute increment rainfall pattern produced by a disaggregation model based on geometric similarity. For 98 test storms the peak discharges produced by the ANN model rainfall pattern had a median under-prediction of 16.6%. This relative error was less than the median under-prediction in peak discharge when using the observed 15-minute rainfall patterns aggregated to hourly increments (40.8%), and when using rainfall patterns produced by the geometric similarity rainfall disaggregation model (21.9%).

Cascade-based disaggregation of continuous rainfall time series: the influence of climate

Abstract. Rainfall data of high temporal resolution are required in a multitude of hydrological applications. In the present paper, a temporal rainfall disaggregation model is applied to convert daily time series into an hourly resolution. The model is based on the principles of random multiplicative cascade processes. Its parameters are dependent on (1) the volume and (2) the position in the rainfall sequence of the time interval with rainfall to be disaggregated. The aim is to compare parameters and performance of the model between two contrasting climates with different rainfall generating mechanisms, a semi-arid tropical (Brazil) and a temperate (United Kingdom) climate. In the range of time scales studied, the scale-invariant assumptions of the model are approximately equally well fulfilled for both climates. The model parameters differ distinctly between climates, reflecting the dominance of convective processes in the Brazilian rainfall and of advective processes associated with frontal passages in the British rainfall. In the British case, the parameters exhibit a slight seasonal variation consistent with the higher frequency of convection during summer. When applied for disaggregation, the model reproduces a range of hourly rainfall characteristics with a high accuracy in both climates. However, the overall model performance is somewhat better for the semi-arid tropical rainfall. In particular, extreme rainfall in the UK is overestimated whereas extreme rainfall in Brazil is well reproduced. Transferability of parameters in time is associated with larger uncertainty in the semi-arid climate due to its higher interannual variability and lower percentage of rainy intervals. For parameter transferability in space, no restrictions are found between the Brazilian stations whereas in the UK regional differences are more pronounced. The overall high accuracy of disaggregated data supports the potential usefulness of the model in hydrological applications. Keywords: Rainfall, temporal disaggregation, random cascade, scaling, semi-arid, temperate climate.

Temporal rainfall disaggregation based on scaling properties

The present study concerns disaggregation of daily rainfall time series into higher resolution. For this purpose, the scaling-based cascade model proposed by Olsson (1998) is employed. This model operates by dividing each rainy time period into halves of equal length and distributing the rainfall volume between the halves. For this distribution three possible cases are defined, and the occurrence probability of each case is empirically estimated. Olsson (1998) showed that the model was applicable between the time scales 1 hour and 1 week for rainfall in southern Sweden. In the present study, a daily seasonal (April-June; 3 years) rainfall time series from the same region was disaggregated by the model to 45-min resolution. The disaggregated data was shown to very well reproduce many fundamental characteristics of the observed 45-min data, e.g., the division between rainy and dry periods, the event structure, and the scaling behavior. The results demonstrate the potential of scaling-based approaches in hydrological applications involving rainfall.