Effects of Heavy Rainfall and Future Flood Intensity

Introduction

Heavy rainfall is commonly used as important climate parameter to determine the changes in global climate change. Referring to the Intergovernmental Panel on Climate Change (IPCC) Fifth Assessment Report (AR5) [1], increment at the rate 0.65-1.06°C of global temperature over period of 1880-2012 was observed. Study on extreme temperature events indicate an existence of warming trend at certain meteorological stations in Malaysia [2]. It was found that increment in sea surface temperature due to climate change has caused changes in rainfall trend and cause extreme rainfall [3]. Changes in climate event, effects of heavy rainfall and extreme weather has been observe since 1950 [1] and with a long period extreme weather can cause unpredicted flood event.

Heavy Rainfall as the Cause of Flood

Heavy rainfall is the main cause of flood occur around the world. It has been recognized among the worst natural disaster in the world causing tremendous damages to properties, environment and losses of life [4]. Approximately more than 100 lives per event were lost from previous incident with no less than 60 extreme flood disasters in 1980 [5]. Large flood occurred mostly in humid tropical areas of Asia and South America. Another cause of flood is modification on the physical environment which could also lead to serious harm to various ecosystem especially river.

Malaysia experienced many floods in the past 50 years whether it is of minor or major flood. The major flood that occurred in January 1971, has resulted in more than RM 200 million losses and 61 deaths throughout Peninsular Malaysia including the capital Kuala Lumpur [7]. In addition, between late December 2006 and early January 2007, Malaysia faced another heavy rainfall event and a major flood at the south of Peninsular Malaysia during the northeast monsoon. The estimate total loss due to this flood was RM 1.5 billion including RM 237.1 million for damages to infrastructure [8]. Another extreme rainfall and flood occurred from 9 to 11 December 2014 which eventually lead to extreme flood over the east coast of Peninsular Malaysia [9]. This flood event had been recorded as one of the worst floods occuring in Malaysia and restoration cost used by Malaysia Government value over RM 1 billion.

With more evident on increasing extreme rainfall coupled with rapid land development that contribute to more impervious areas, future floods are expected to be more extreme and occur more regularly [11]. Due to these events, rainfall possibly shows changes in trend and affects previous study. However, the historical rainfall data with no climate effects are insufficient in providing reliable information of the future rainfall behaviour. These proven by the list of extreme events occur around Malaysia. Meanwhile, warmer climate likely will increase number of heavy rainfall and will produce more intense events. This could lead to longer dry, lower wet days with intense rainfall and high risk of floods. Thus, flood events caused by enhanced rainfall intensity seem to be a major concern soon. This occurrence can contribute to severe and diverse impact such as losses of human lives, destruction to properties and infrastructures, losses of livestock and disruption to the ecosystem and the environment at large.

Rainfall trend change by time will affect the understanding of present rainfall behaviour. Rainfall trend is important in various economic and infrastructures study such as efficiency in design of dams, waterway, highway, urban storm sewer, farm terrace and many other hydraulic structures design with consideration of rainfall variable. Therefore, in this study, rainfall at a specific area was analysed to determine the any significant changes in its trend. The result will provide facts to the importance of considering climate change in frequency analysis.

Frequency analysis is the method commonly used to estimate the intensity of rainfall that would occur for a certain return period. Rainfall event that is often analysed by using past observation data with simplified assumptions might not be able to accurately assess the properties of the climate change in the future. Recent literatures have found that process based Generalize Circulation Models (GCM) are the essential devices used to make projections of future climate under scenarios of anthropogenic greenhouse gas effects [12]. Climate change information provided by GCMs model used to give a more realistic result for future frequency analysis. Thus, gave better information to design future hydraulics structure and modelling future hydrological conditions.

Problem Statement

Flood event affecting most state on the east coast of Peninsular Malaysia was due to the northeast monsoon which normally occurs on October to March. On December 2014, a total of 3,930 people evacuated from their home in several areas of Kelantan due to the days of heavy rain. Rainfall depth recorded for continuous three days on 21st to 23rd December 2014 are equivalent to the normal recorded rainfall depth for 64 days. Abnormally on the rainfall intensity have led to extreme flood and caused water flow to exceed the capacity of river bank. This extreme event had caused tensive and severe damages to properties, public facilities, local infrastructures and the environment. The three main rivers of Kelantan consist of Sungai Galas, Sungai Lebir and Sungai Kelantan rose drastically above normal safe level.

Extreme rainfall and flood are expected to recur in the future. Researchers and modelers had critically examined and improve frequency analysis to in view of the more frequent intense rainfall in recent years. There is a pressing need now to develop future Intensity Duration Frequency (IDF) by considering the increasing occurrence of extreme rainfall in order to maintain the accuracy of frequency analysis method for current and future rainfalls. This requires future rainfall data that can be simulated from GCM.

Heavy rainfalls may aggravate flood problems and causing much bigger damage compared to with normal flood. The construction of future IDF is critical to predict the possibility of flood occurrences, formulate emergency action plan (EAP) and determining the appropriate insurance premium. In addition, the rainfall values for various return periods that can be derived from IDF curves are useful as inputs in flood modelling and designing hydraulic structures.

Objectives of the Study

This study is concerned with developing future Intensity Duration Frequency Curve by incorporating the effects of climate changes. In so doing, the following specific objectives are outlined:

1. To determine the trend of annual maximum rainfall event using historical data.
2. To bias-correct simulated CCSM3 rainfall data
3. To construct future IDF curves for selected stations in Kelantan.

Scope of the Study

The first objective in this study is analysing the rainfall trend using Mann-Kendall Trend test and Sen’s Slope analysis. The data used for trend analysis is in annual maximum series of hourly basis. This type of data is more accurate since Department of Irrigation and Drainage in Malaysia used tipping bucket of hourly scale. There might be error if the data used in daily instead of hourly. Trend analysis was used to understand the behaviour of past rainfall pattern. The increment or decrement in intensity rainfall can be concluded through the analysis.

The second objective is to bias correct GCM data. GCMs model compute simulated data to understand the behaviour of rainfall due to climate change. The simulated data may contain error and bias. A quantile mapping is one of commonly used bias correction method to correct data by understanding the statistical properties of the data itself. A simple and detail methodology is provided in this study and can be used for most GCMs model.

The final objective of this study is producing a future IDF. To construct an IDF we applied distribution analysis to calculate the return period for different duration. Return period used in this study are 2-, 5-, 10-, 50- and 100-year. The durations used are in calculation of moving total for 1-, 2-, 3-, 4-, 5-, 6- and 7-hour. This study used two types of data consist of observe data and GCM data.

Rainfall trend and Intensity Duration Frequency (IDF) is employed to give the result of this study. Then, the IDF graph is plotted with different return period and compared using the graphical analysis. Climate change factor (CCF) is calculated to represent as climate index and be applied to IDF to produce more relevant result with implication of predicted climate change.

Importance of Study

The greatest need of information relating to rainfall analysis is apparently for various kind of area such as economic, infrastructures, planning and management. These areas worked with consideration of the frequency of certain rainfall events. In enhancing the rainfall analysis and its application, scientist and engineer are therefore require possessing a working knowledge of the rainfall analysis with respect to its principal and procedure.

The existing analyses are numerous, but the effectiveness can be enhanced by additional information and analysis. Trend analysis needed to determine if there was a change in the past. With a significant result, it will show the climate change in the past and by assumption, it can possibly be change from in future. Therefore, it is important to include climate change in the rainfall analysis to increase accuracy.

To use climate change as a factor, simulated CCSM3 is used. Unfortunately, it cannot be used directly to do analysis. In this study, we applied quantile mapping and application of using distribution for extrapolating data to bias correct rainfall simulated CCSM3. This outcome in the transfer function which is unique for every station can be applied to future data. Finally, after bias correcting the data, simulated CCSM3 are ready to be analyse.

The result of this study is producing a future IDF. The differences between this study and other study in Malaysia in constructing IDF are the data used and implementation of climate change. Hydrological modelling such as flood model uses hourly or smaller rainfall period to model a better and accurate simulation. IDF play as the main role to provide the input of design rainfall data in hourly duration. Therefore, to compute a small duration prediction, it is needed to use small data duration. The implementation of climate change will be an advantage to increase the accuracy for future analysis.

Conclusion

In conclusion, this study provides knowledge and understanding of the trend analysis, use of GCMs model and application of IDF curve. Thus, throughout this study, the result are trend analysis which can describe the past behaviour of rainfall in Kelantan and the information of intensity, duration and frequency which can be used as input for flood modelling. Therefore, local authorities could have a better planning for flood mitigation and flood evacuation strategy.

Works Cited

1. Intergovernmental Panel on Climate Change (IPCC). 'Fifth Assessment Report (AR5).' IPCC, 2014.
2. Kwan, M. P., & Wong, M. S. 'Climate Change and Air Quality in Urban Areas: A Review.' Climatic Change, vol. 125, no. 4, 2014, pp. 469-482.
3. Tan, P. Y., & Gan, T. Y. 'Trends and Variability in Rainfall in Malaysia.' Water Resources Management, vol. 27, no. 11, 2013, pp. 3989-4006.
4. Ahern, M., Kovats, R. S., Wilkinson, P., Few, R., & Matthies, F. 'Global Health Impacts of Floods: Epidemiologic Evidence.' Epidemiologic Reviews, vol. 27, 2005, pp. 36-46.
5. National Oceanic and Atmospheric Administration (NOAA). 'Natural Hazard Statistics.' NOAA, 2021.
6. Ahmad, Z. 'Floods in Malaysia: A Study on Their Causes and Extent.' Geografia-Malaysian Journal of Society and Space, vol. 5, no. 3, 2009, pp. 42-53.
7. Department of Irrigation and Drainage (DID) Malaysia. 'Flood Management in Malaysia.' DID Malaysia, 2013.
8. Ho, C. H., & Saifullah, A. M. 'Hydro-Climatic Trends of the Johor River Basin, Malaysia.' Journal of Hydrology, vol. 529, 2015, pp. 1627-1640.
9. Kwan, C. M., & Yusop, Z. 'Rainfall Trends and Variation Analysis in Peninsular Malaysia.' Sains Malaysiana, vol. 40, no. 4, 2011, pp. 401-408.
10. Gao, X., & Pal, J. S. 'Mean, Variability, and Trends in Northern Hemisphere Snowfall.' Journal of Climate, vol. 17, no. 19, 2004, pp. 3661-3675.
11. Chu, J. T., & Xia, J. L. 'Rainfall Frequency Analysis for Future Periods in Connecticut.' Journal of Hydrologic Engineering, vol. 14, no. 9, 2009, pp. 1005-1010.
12. Wilby, R. L., & Wigley, T. M. 'Downscaling General Circulation Model Output: A Review of Methods and Limitations.' Progress in Physical Geography, vol. 21, no. 4, 1997, pp. 530-548.