Mitigating flood risk at Shah Alam, Malaysia for sustainable development

Abstract

The natural environment and infrastructure are severely damaged by floods, which are becoming frequent occurrences. It may not be able to stop floods given the impending climate change and variability of the weather. Yet mapping flood risk can help with both mitigation and prevention of flooding. The Shah Alam experiences flash floods yearly, which is why it was chosen. An interactive approach to decision-making under multi-criteria decision analysis (MCDA), the analytical hierarchy process (AHP), is used. The flood risk zone is divided into five categories very low-risk (32.38%), low-risk (22.25%), moderate-risk (23.17%), high-risk (18.87%), and very high-risk (3.33%) respectively. Based on the data gathered for this study, the economic losses are identified, low-risk areas (RM 2091), moderate-risk areas (RM 2343), and high-risk areas (RM 2728). The historical flood events occur in the Seksyen 36, 18, 17, and most of the southern part of Shah Alam, Malaysia. Rainfall and the Klang River have the biggest effects on these regions, with water drainage systems being critical in the Shah Alam area to reduce the risk of floods. When analysing flood risks in the Shah Alam region and throughout the nation, important stakeholders both state and non-state actors can benefit greatly from the flood susceptibility map provided by this study.

Introduction

Floods were thought to be between maximum destructive natural catastrophes, and causation of enormous damage, out of many kinds of dangers [1]. Worldwide property and human casualties are caused by this occurrence every year. One nation where flooding has a detrimental effect is Malaysia, which has several states affected, including Terengganu and Kelantan [2]. Reports now in circulation indicate that forty percent of all damages in Malaysia are caused by flooding. According to the Department of Irrigation and Drainage (DID), floods can occur in 9% of Malaysia’s total land area, with damages estimated to be around USD 0.3 billion annually [3]. It is anticipated that the frequency of this calamity will rise as a result of increased urbanization and unplanned development, deforestation, and precipitation brought on by climate change in vulnerable areas [4]. Given the detrimental effects of floods, identifying vulnerable places is crucial [1]. It is well-established that measuring flood susceptibility is a crucial first step in controlling and preventing future floods [5, 6]. Predicting floods accurately is challenging due to their intricate characteristics [7].

The primary global source of irreversible harm is natural disasters [8]. Flood events are annually welcomed in Malaysia, primarily during the monsoon seasons [9]. Traffic, communities, farming, and means of subsistence have all suffered significant losses as a result of these floods [10]. Such effects can be reduced by doing an appropriate and accurate susceptibility examination. As such, one of the main objectives of governments and scientists is to produce realistic flood models. In the literature over the years, numerous hydrological techniques have been employed [11]. A thorough examination of rivers and inundation zones cannot be accomplished by applying traditional hydrological approaches, like data-driven approaches and physical-based rainfall-runoff modeling techniques [12]. The explanation is that, although the morphology of the river is not stable and has dynamic characteristics due to the strong erosive potential, the hydrological approaches follow a one-dimensional procedure. Due to Malaysia’s high frequency of flooding, this natural disaster is the most dangerous one, resulting in numerous fatalities, property losses, and ecological damage [13]. Numerous accounts of floods in Malaysia have been kept on file since the 1920s. According to the DID, floods can harm 9% of Malaysia’s landform zone (29,800 km²) and 22% of its population (4.82 million) as a result [13]. Over the past few decades, Kelantan, one of Malaysia’s 13 states, has been severely impacted by the yearly monsoon floods. Heavy monsoon rainfall has recently caused floods in Malaysia, posing a particular threat to states along the country’s east coast, including Kelantan, Terengganu, and Pahang [14].

Maps of flood susceptibility have become more frequently created using Geographic Information System GIS) thanks to the quick availability of satellite remote sensing (RS) data and the advancement of technology [15, 16]. Moreover, GIS is a helpful tool for looking at multi-dimensional occurrences like floods. One of the greatest extensively applied methods aimed at implementing a multi-criteria-based decision support system to maximize decision-making under a collection of qualitative, quantitative, and occasionally competing elements is the Analytic Hierarchy Process (AHP) [17, 18]. According to a process suggested by Saaty [19], the best answer is found by performing a pairwise comparison after the variables are ordered in a hierarchy. For several tasks, including mapping flood risk, combining AHP and GIS has proven advantageous [20]. To arrive at the best choice, socioeconomic, environmental, and technical goals could all be combined using the Multi-Criteria Decision Analysis (MCDA) techniques [21]. Because complex decision issues frequently involve irreconcilable incommensurable criteria, the MCDA has been generally acknowledged as an important tool for analysis [22]. One qualitative MCDA technique, the AHP, depends on the expertise of the expert in weight assignment for both the process and its implementation [23]. In this analysis, flood risk assessment is conducted in the Shah Alam area, Malaysia with MCDA and AHP techniques where three main criteria environmental, economic, and social criteria were applied to detect the flood risk zone in the Shah Alam area. The weighted values are generated by using the perspectives of twenty interviewees from various sectors, including research, academia, and stakeholders. This study produces weighted value averages based on the values. This study's results can help to prepare a flood mitigation strategy, novel adaptation techniques, awareness, and disaster risk reduction towards sustainable development in the Shah Alam area.

Study area

The research area includes the area around the National Botanic Gardens Shah Alam (3°N Latitude and 101°E Longitude) and the area that was originally known as “Bukit Cahaya Sri Alam Agricultural Park”. The National Botanic Gardens Shah Alam is a protected area and forest reserve that is recognized as an urban forest. During the northeast monsoon season, which runs from October to March in Shah Alam, the city receives sunny, wet, hot, humid tropical weather (Fig. 1). The expected range for temperatures is 22 °C to 23.5 °C for minimum values and 31 °C to 33 °C for highest values. The annual rainfall is approximately 3300 mm, and the relative humidity ranges from 72 to 78%. The study area's northernmost point is the National Botanic Garden, while Shah Alam’s southernmost section is traversed by the Klang River, which spans 304 km². Due to the intense rains, the Klang River and its environs frequently flood, which negatively affects local livelihood and causes significant financial losses. (https://www.water.gov.my/).

Fig. 1 [Images not available. See PDF.]

Study area map of Shah Alam, Malaysia

Methods and materials

Applied datasets

The rainfall data was collected from 2013 – 2022 in Shah Alam, Malaysia for flood analysis (Table 1). Based on the station rainfall data, the distribution of rainfall is calculated through the inversed distance weighted (IDW) in GIS software. The JPS offers two rainfall stations, located at UiTM Shah Alam and Jalan Acob Estate. The Jalan Acob Estate has an average dataset of 3.018 mm, while UiTM Shah Alam has an average of 205.1558333 mm recorded. The GIS software will be used to map the data once the average rainfall for each area has been obtained. The rainfall region in the map will be visualized using the data ‘weighted overlay’ method, which was computed on the examining area layer in the GIS using Excel. Three sorts of maps environmental, social, and economic have been developed for this study through the process of mapping flood risk (Fig. 2).

Table 1. Details about the datasets

Datasets	Data source	Resolution	Website
Rainfall	Department of irrigation and drainage	N/A	https://www.water.gov.my/
Distance to river	SRTM DEM	30 m	https://earthexplorer.usgs.gov/
Drainage density		30
Elevation		30 m
Slope		30 m
LULC	Landsat 8	30 m
Distance to road	BBBike, google earth	N/A	https://extract.bbbike.org/
Population density	Majlis Bandaraya Shah Alam (MBSA)	N/A	https://www.malaysia.gov.my/portal/subcategory/208?language=my
Economic losses	Household survey

Fig. 2 [Images not available. See PDF.]

Adopted methodology to calculate the flood risk in Shah Alam, Malaysia

Criteria

Rainfall is the mean data rainfall at Shah Alam from 2013 to 2022. In this study, the rainfall component is crucial since it results in river overflow, which in turn results in flooding (Fig. 3). The station rainfall data is distributed in GIS software through the Inverse Distance Weighted (IDW) interpolation method. By estimating the value z at position x using a weighted mean of neighbouring data, an IDW interpolation is performed. As a result, deep rainfall leads to major flooding, while shallow rainfall causes mild flooding [24]. Rainfall was classified into five classes: very low class (scaling 1 to 94.6–71.6 mm), low class (scaling 2–118 mm), moderate class (scaling 3–141 mm), high class (scaling 4–164 mm), and very high class (scaling 5 to 165–187 mm).

Fig. 3 [Images not available. See PDF.]

Criteria Map for the flood risk analysis in Shah Alam

Distance from the river network component was significant in defining the flood-related risk zones. Based on a study by Fernández, 2010, the risk of flooding is highest in areas close to river networks, and it becomes less harmful the further one flows from the riverbed. The distance from the river is calculated in GIS software with river data and a ‘multi-ring buffer’ tool within the Shah Alam area. Based on Fig. 3, there was the classification of distance to the river had 5 very high categories (1 km), high (3 km), moderate (7 km), low (13 km), and very low (20 km).

Drainage density for the Klang River and Damansara River in the Shah Alam area. Flooding is mostly caused by factors such as drainage density. When the drainage density is large, the water runoff rate becomes critical [25]. The river datasets are applied for mapping drainage density in Shah Alam through the “kernel density” tool in GIS software. The drainage density map in this study was separated into five groups. Low flood risk (60.2–120 km), moderate flood risk (121–180 km), high flood risk (181–240 km), very high flood risk (241–301 km), and very low flood risk (0–60.1 km) are all characterized by a lower drain-age density that is less affected by flooding.

Elevation at Malaysia, Shah Alam. Flooding occurs more frequently in lowland areas when rainwater flows from higher elevations to lower elevations [26]. In a GIS, this calls for an SRTM DEM for raster-based hydrological analysis (Fig. 3). Extremely low risk was assigned to the regions with very high elevations between 114 and 229 km. Low-risk individuals were those whose heights were between 70.1 and 113 km. Moderate risk was assigned to regions with elevations between 42.1 and 70 km.

Slope map at Malaysia, Shah Alam. The slope is thought to be yet another important element that could cause a flood [27]. The slope affects surface runoff velocity and volume as well as groundwater infiltration, which is calculated in GIS software through the “slope” tool. The degree of slope was classified in this study as follows: low flood risk, scaled 4 (11.8–18%), moderate flood risk, scaled 3 (7.05–11.7%), very high flood risk, scaled 1 (0–3.52%), and very low flood risk, scaled 5 (18.1–49.9%).

LULC at Shah Alam represents the fundamental characteristics of the Earth system, deeply entwined with a variety of human pursuits and the natural environment (Fig. 3). The LULC classification is prepared from Landsat 8 data using supervised classification with a maximum likelihood algorithm in GIS software [28]. In this research, LULC has five categories to investigate areas which are low risk (barren land), moderate risk (build-up area), high risk (vegetation), and very high risk (water body) will affect the flood at Shah Alam.

Distance to the road the distance of each location from the road on the Shah Alam map is a significant aspect in identifying whether or not a region is prone to flooding. Consequently, low-intensity rainfall overwhelms regions with a densely packed road system, resulting in flooding [24]. The classification of this research is as follows: low flood risk scaled 4 (10 km), moderate flood risk scaled 3 (5 km), very low flood risk scaled 5 (> 15 km), and very high flood risk scaled 1 (1 km).

Population density is obtained from the Shah Alam City Council (MBSA) report, where section-wise population density is applied to calculate the flood-related risk zone identification in the Shah Alam region. Most of the people are affected because of the high frequency of floods in this region. Section-wise population density is prepared from population data in Shah Alam.

Economic loss is divided into three categories, based on the household survey in the Shah Alam area. Based on the household survey, economic losses are identified as high, moderate, and low loss zones, where the southern part of Shah Alam has a high economic loss area due to the highly frequent flood. The northern part has fewer economic losses due to the low flood effect in this area.

MCDA and AHP methods

AHP is one of the models that GIS-MCDA approaches use the most commonly [29]. This method simplifies the weighting procedure by using pairwise comparison matrices, which arrange expert knowledge. AHP is a user-friendly method for decision-making in complex, unstructured, and multi-attribute problems. [30]. An extensive number of natural hazards, like earthquakes, wildfires, floods, and landslides, have been evaluated using the AHP model. [31]. Experts’ in-depth knowledge is required to generate valuable pairwise comparison matrices (PCM) [32]. The AHP makes use of PCM’s unique qualities, which are theoretically reciprocal, consistent, and quadratic. It’s advised to use an underlying scale with values ranging from 1 to 9 for paired comparisons in the AHP model for expert evaluation [30]. The concept of transitivity, which is defined as follows, is the central idea of AHP: if B1 > B2 and B2 > B3, then B1 > B3. Along with the other three parts, this also applies to B1, B2, and B3. Most of the conditioning factor weightings in AHP stem from the transitiveness concept. A consistent PCM method would need to demonstrate that if 2B1 > B2 (i.e., B1 was twice as wanted as the value B2), and 4B2 > B3, then 8B1 > B3. The transitivity principle explains this. Subsequently, consistency must be consistently confirmed throughout the AHP process, which could involve. where n is the total number of criteria and CI is the maximum computed matrix eigenvector we ${\mathrm{\lambda }}_{\mathit{max}}$. The mean consistency index value that depends on the matrix instruction that is presumably through Saaty in 1997 is known as the randomness index (RI) (Tables 2, 3 and 4).

Table 2. The pairwise comparison matrix for the multi-criteria decision analysis (MCDA)

Factor	Rainfall	Distance to river	Drainage density	Elevation	Slope	LULC	Distance to road	Population density	Economic losses
Rainfall	1	2	2	3	2	1	4	6	5
Distance to river	0.50	1	2	3	5	2	1	3	5
Drainage density	0.50	0.50	1	5	3	1	4	2	3
Elevation	0.33	0.33	0.2	1	3	2	1	1	5
Slope	0.50	0.20	0.33	0.33	1	1	1	2	4
LULC	1.00	0.50	1	0.5	1	1	1	1	2
Distance to road	0.25	1.00	0.25	0.33	1	1	1	1	3
Population density	0.17	0.33	0.5	1	0.50	1	1	1	2
Economic losses	0.20	0.2	0.33	0.33	0.50	0.5	0.33	0.50	1
Total	4.45	6.06	7.61667	14.50	17.00	10.5	14.33	17.50	30

Table 3. The standardized matrix and the weighted dissemination for the multi-criteria decision analysis (MCDA)

Factor	Rainfall	Distance to river	Drainage density	Elevation	Slope	LULC	Distance to road	Population density	Economic losses	Criteria weight	Criteria Weight (Rounded)
Rainfall	0.22	0.33	0.26	0.21	0.12	0.10	0.28	0.34	0.17	0.23	23
Distance to river	0.11	0.16	0.26	0.21	0.29	0.19	0.07	0.17	0.17	0.18	18
Drainage density	0.11	0.08	0.13	0.34	0.18	0.10	0.28	0.11	0.10	0.16	16
Elevation	0.07	0.05	0.03	0.07	0.18	0.19	0.07	0.06	0.17	0.10	10
Slope	0.11	0.03	0.04	0.02	0.06	0.10	0.07	0.11	0.13	0.08	8
LULC	0.22	0.08	0.13	0.03	0.06	0.10	0.07	0.06	0.07	0.09	9
Distance to road	0.06	0.16	0.03	0.02	0.06	0.10	0.07	0.06	0.10	0.07	7
Population density	0.04	0.05	0.07	0.07	0.03	0.10	0.07	0.06	0.07	0.06	6
Economic losses	0.04	0.03	0.04	0.02	0.03	0.05	0.02	0.03	0.03	0.03	3
Total	1	1	1	1	1	1	1	1	1	1.00	100

Table 4. Estimation of the consistency ratio

Factor	Rainfall	Distance to river	Drainage density	Elevation	Slope	LULC	Distance to road	Population density	Economic losses	Consistency vector
Rainfall	0.23	0.36	0.32	0.30	0.15	0.09	0.29	0.36	0.17	10.10
Distance to river	0.11	0.18	0.32	0.30	0.38	0.18	0.07	0.18	0.17	10.41
Drainage density	0.11	0.09	0.16	0.49	0.23	0.09	0.29	0.12	0.10	10.59
Elevation	0.08	0.06	0.03	0.10	0.23	0.18	0.07	0.06	0.17	9.96
Slope	0.11	0.04	0.05	0.03	0.08	0.09	0.07	0.12	0.14	9.65
LULC	0.23	0.09	0.16	0.05	0.08	0.09	0.07	0.06	0.07	9.80
Distance to road	0.06	0.18	0.04	0.03	0.08	0.09	0.07	0.06	0.10	9.77
Population density	0.04	0.06	0.08	0.10	0.04	0.09	0.07	0.06	0.07	10.03
Economic losses	0.05	0.04	0.05	0.03	0.04	0.05	0.02	0.03	0.03	9.96

1

$$CI=\frac{\sum ({\mathrm{\lambda }}_{\mathit{max}}-n)}{(n-1)}$$

Based on professional judgment or expertise, the paired comparison matrix's modified rational illogicality was determined was gradually classified using this equation, and the results showed that the AHP was predictable by CR. Values of CR values < 0.1 were the dependable pairwise matrix sign. The RI values were redefined and changed when a criterion was applied.

2

$$CR=\left(\frac{\mathit{CI}}{\mathit{RI}}\right)$$

RI is the abbreviation for random index. For the randomly generated pairwise comparison matrix, where n = 2, 3, 4, 5, 6, 7, 8, and 9, the consistent index is denoted by CI and the random index by RI (Tables 5, 6). When the consistency level is less than 0.10, it indicates an adequate consistency level; when the consistency level is more than 0.10, it indicates an undesirable consistency level. Based on the AHP calculation, the CR value identified $0.0889$, which is under acceptable limits based on Saaty’s analysis.

Table 5. Random Consistency Index (RI) (Saaty, 1977, 1990)

n	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
Rl	0	0	0.58	0.9	0.12	1.24	1.32	1.41	1.45	1.49	1.51	1.48	1.56	1.57	1.59

Estimation of lambda (λ) = {(10.10 + 10.41 + 10.59 + 9.96 + 9.65 + 9.80 + 9.77 + 10.03 + 9.96)/9} = 10.031

Estimation of Consistency Index:$CI=\frac{(\mathit{\lambda }-\mathit{n})}{(\mathit{n}-1)}$

Where, λ = Lambda $CI=\frac{(10.031-9)}{(9-1)}$

n = Number of Criteria $CI=0.1289$

Estimation of Consistency Ratio $CR=\left(\frac{\mathit{C}\mathit{I}}{\mathit{R}\mathit{I}}\right)$

$CR=\frac{0.1289}{1.45}$

$CR=0.0889$

Table 6. Criteria table for flood risk assessment

Flood risk criteria	Unit	Class	Risk class ranges and ratings	Risk class rating	Weight (%)
Environmental criteria
		71.6–94.6	Very low	1	23
Rainfall	mm	94.7–118	Low	2
		119–141	Moderate	3
		142–164	High	4
		165–187	Very high	5
		1	Very high	5	18
Distance to river	km	3	High	4
		7	Moderate	3
		13	Low	2
		> 20	Very low	1
		0–60.1	Very low	1	16
Drainage density	km	60.2–120	Low	2
		121–180	Moderate	3
		181–240	High	4
		241–301	Very high	5
		0–21	Very high	5	10
Elevation	m	21.1–42	High	4
		42.1–70	Moderate	3
		70.1–113	Low	2
		114–229	Very low	1
		0–3.52	Very high	5	8
Slope	%	3.53–7.04	High	4
		7.05–11.7	Moderate	3
		11.8–18	Low	2
		18.1–49.9	Very low	1
		Barren Land	High	5	9
LULC	Level	Build up Area	Very high	7
		Vegetation	Moderate	3
		Water Body	Low	2
Distance to road	km	3	High	4	7
		5	Moderate	3
		10	Low	2
		> 15	Very LOw	1
Social criteria
		1000–7440	Low	1	6
Population density	Person	7441–18,187	Moderate	2
		18,188–43,791	High	3
		43,792–86,548	Very high	4
Economic criteria
Losses and damages	Ringgit (RM)	2091	Low	1	3
		2343	Moderate	2
		2728	High	3

Furthermore, the second map shows the Shah Alam, Malaysia's distance from the river. Drawing along the Klang and Damansara rivers in Google Earth Pro v7.3 will be the first step in creating the river raster. The multiple-ring buffer method will be used to provide a visual representation of the distance from the river once the river raster has been loaded into GIS software. A visual representation of the distance along the river can be obtained by entering five distinct distances: 1, 3, 7, 15, and 20 km. This map's methodology is identical to that of the road's distance. Using GIS software's multiple ring buffers and the road raster from Google Earth Pro, one can visualize five distinct distances in km by converting the format to KML.

Hotspot analysis

Hotspot analysis is usually used to investigate heavily damaged places where real datasets are more important. In other cases, hotspot areas are not examined in road accident or safety identification studies [33]. Hotspots were used to create equitable spacing between the establishment of flood-prone zones that are very low, low, medium, high, and extremely high [34]. Aside from manually identifying the hotspot area, many software applications apply the polygon, line, and point datasets for hotspot inspection.

3

$$G_i^{\ast }(d)=\frac{{\sum }_{j=1}^{\text{n}}{W}_{\mathit{ij}}\left(d\right)x_j-w_i^{\ast }\overline{x}}{s\left[\frac{W_i^{\ast }(\text{n}-W_i^{\ast }}{\text{n}-1}\right]_{}^{\frac{1}{2}}}$$where x is the probability of the spatial data; $W_{\mathit{ij}}$ denotes the spatial weightage of the chosen datasets, which ranges from $j$ to $i$; $W_{\mathit{ij}}$ was used to identify the sum weightage similar to, $W_i$; s is the standard deviation of the x values, and, $\overline{x}$. shows the mean value of the chosen datasets.

Software use

The GIS software version is applied for the criteria map generation of the manuscript. Different tools in GIS like multiplying buffer, slope, kernel density, and finally weighted overlay are applied for flood risk assessment. Utilizing Microsoft Excel, the Pairwise comparison matrix is generated. Google Earth Pro software version 7.3 is used for river and road digitization.

Validation through ROC

The receive operating characteristics (ROC), to find the best model, in flood risk analysis. “Area Under Curve” attitudes are represented by the ROC or AUC. The AUC method covers the whole two-dimensional area from the beginning of the ROC curve to its end. A receiver operating characteristic curve, often known as a ROC curve, is a type of visual figure that shows how a binary classifier organization can analyse data given a variety of judgment thresholds.

Result and discussion

Flood Susceptibility of Shah Alam

The result of integrating flood risks and vulnerabilities at a particular place is flood risk. Therefore, assessing, gathering, and systematically analysing variables is necessary. Figure 4 shows the flood risk estimations for the Shah Alam, Malaysia. The study weighted each subject indicator class's and layer’s relevance and contribution to risk and susceptibility. The locations classified for flood risk were determined by merging the two factors after the indicators were assessed for hazard and susceptibility using the ‘weighted overlay’ technique. The raster calculator found in the spatial analyst tools was used for the entire GIS process (Fig. 4). These are the next major elements that significantly influence the classification of flood-prone areas. The procedure of producing the flood susceptibility map for the study region was not significantly impacted by any of these variables. Based on the literature and expert analysis, the flood risk zone is divided into five categories very low-risk zone (98.43 km²), low risk-zone (67.65 km²), moderate risk zone (70.43 km²), high-risk zone (57.36 km²), and very high-risk zone (10.13 km²) respectively (Table 7). The southern is notified very high to high flood risk zone in Shah Alam, Malaysia. Because of the Kallang River and precipitation, these regions are the most affected land in the Shah Alam area where the water drainage system is essential for flood risk reduction. Based on the AHP technique, the southern and south-eastern regions of the northern zone in Shah Alam have high potential zones, while the areas along the Kallang River have very highly sensitive zones. Southerly Shah Alam is more vulnerable than its northern counterpart due to heavy precipitation and streamflow issues. Based on the AHP flood susceptibility analysis, the northern portion is more vulnerable in U8, U13, Sect. 13, U6, U14, and U3. Shah Alam’s most vulnerable areas are found in the southern region, specifically in parts 32, 28, 34, 36, 18, and so on, depending on the population density in those parts. The livelihood in this area is impacted by flood threats in addition to the environmental impact. Based on the previous flood data, the current flood risk map is validated through the ROC curve, and the curve value is observed at 0.886, which is an acceptable limit of validation (Fig. 4b).

Fig. 4 [Images not available. See PDF.]

a Flood risk map of Shah Alam, Malaysia. b ROC curve for validation of the flood risk map in Shah Alam using previous flood data

Table 7. Flood risk zone area of Shah Alam, Malaysia

Flood risk zone	AHP method
	Area (Km2)	Area (%)
Very high-risk zone	10.13	3.33
High risk zone	57.36	18.87
Moderate risk zone	70.43	23.17
Low risk zone	67.65	22.25
Very low risk zone	98.43	32.38

One important aspect that raises the area's level of flood susceptibility is LULC growth in low-lying areas. Examples of land use development include building roads and structures. For this reason, there is an increasing risk of flooding in this area. [2]. Next, point-based datasets are generated with the locations, corresponding years, coordinates, and number of incidents. These datasets are then obtained as a shapefile that shows the flood areas and is used to verify the models. 2009 flood events impacted 225 locations in total, with Sungai Pinang (near Jalan P. Ramlee) suffering the greatest damage with 192 flood events [35].

Affected population

This study's social factor is Shah Alam, Malaysia’s Population Density Map. The Population Density Map displays the entire population in 2020 to map social aspects. The Shah Alam City Council (MBSA) report aimed at that year provided the population data. The population data in Shah Alam, Malaysia is segmented based on the respective section. Since Shah Alam is divided into 56 sectors, an Excel spreadsheet containing 56 distinct sets of data will be created. Susceptibility and population density are closely related because more people are exposed to hazardous events in places with high population densities [36]. Five categories have been established for the use of GIS in this study: low (population range of 1000–7440), moderate (population range of 7441–18,187), high (18,188–43,791 population), and extremely high (population range of 43,792–86,548) (Fig. 5a). Additionally, this map displays a few locations that saw flooding in the Shah Alam region between 2017 and 2022.

Fig. 5 [Images not available. See PDF.]

Impact of flood on population and economy, a Population density; b Losses and Damages Area Map

(Source: Household Surveys)

The most common natural disaster in Malaysia is flash floods, which are especially common in the eastern part of Peninsular Malaysia, where heavy rains from November to January almost usually cause substantial damage. Between 1956 and 2007, floods claimed the lives of up to 2.7 million people, or 9% of Malaysia’s total land area (29,000 km²) and 23% of its total urban area. (https://www.water.gov.my/). Without adequate development management measures, floods caused by climate change could cost the country RM100 million a year (https://www.nres.gov.my/ms-my/Pages/default.aspx). Malaysia follows the “precautionary principle” and “no regret policy” when it comes to mitigating and adapting to climate change. The National Climate Committee was established in January 1995 by representatives from various government departments and agencies, businesses, and civil society organizations [14]. The National Climate Change Policy was unveiled in 2009. It is projected that the National Physical Planning Council, headed by Malaysia’s prime minister, will oversee urban development planning activities throughout the country. The council was founded in 2001. The council only discusses the “strict controlling of land development in highlands to safeguard human safety and environmental quality,” as detailed in policy number 21, and does not give enough attention to the threat that climate change-related disasters pose to sustainable urban development. The director general of town and country planning is expected to assist the council's efforts to improve the town and country planning process to achieve sustainable urban development [37].

Economic losses

The study's economic component. Map showing damages and losses at Shah Alam, Malaysia. This approach is also employed in the same manner as the economic component, which is the mapping of losses and damages into three categories: low-risk, moderate-risk, and high-risk locations. Values for losses and damages have been gathered for each location from household surveys carried out within the study area. The number of floods that occurred in the area of Shah Alam, Malaysia, between 2017 and 2022 will be superimposed on the map, which is why these two views are different from one another. Adaptation cannot stop all losses and consequences (Fig. 5b). Based on the data gathered for this study, this map has been classified into three categories: low-risk areas (RM 2091), moderate-risk areas (RM 2343), and high-risk areas (RM 2728), which are places that are very susceptible to floods. This map displays some regions that have seen flooding in the Shah Alam area between 2017 and 2022, in addition to losses and damages.

Hotspot analysis

Hotspot identification of this region is essential for understanding the historical flood variation and confidence region. The flood-based hotspot was calculated in the GIS platform (Fig. 6). That information indicates that the region has serious issues in flood variation mainly in the southern region, mainly in a radius around 2 to 3 km from Klang River. The hotspot map indicates hotspot (99% of confidence level) varies in southern part, like seksyen 36, seksyen 17, seksyen 18, seksyen 16, Taman Seri Muda, Kemuning Permai, and Taman Seri Andalas.

Fig. 6 [Images not available. See PDF.]

Hotspot and IDW generation of historical flood data in Shah Alam

Natural disasters, which include regular flash floods, seriously harm society and are especially concerning when they strike densely inhabited areas and places with a high concentration of economic activity [38, 39]. One such potential economic zone is Kajang Town, which is in Malaysia's District of Kajang. The frequency of flash floods in this area has been rising over time. This area reportedly saw flash floods once in 2002, twice in 2008 and 2011, and three times in 2014 [40]. Two-thirds of Kajang City was inundated by the 0.5–1.5 m high flash flood that occurred in 2008 [41]. Severe flooding caused by basin-wide floods on major river systems and localized flash floods in numerous areas of Malaysia has occurred in recent years (Fig. 7). The most severely affected river basins were the Juru River Basin in Penang, the Pahang River Basin in Pahang, the Setiu River Basin in Terengganu, and the Kinta River Basin in Perak. Comparable floods occurred in the Malaysian state of Perlis in the north in 2005, 2010, and 2011 [42]. According to reports, Segamat, Johor, saw its worst flood in the preceding century on January 12, 2007 [43]. The economic status of Malaysia is significantly impacted by floods as well. For example, the 2000 floods severely affected roughly 2.5 million urban inhabitants since their towns were located on Malaysia's flood plains [44].

Fig. 7 [Images not available. See PDF.]

Flood in Shah Alam, Malaysia

(Source: The Star, Malaysiakini, Reuters)

Therefore, the actual research placed equal importance on the physical losses and damages caused by floods. Regarding the building structure, the estimated loss resulting from building contents was 41%, but the cost of new material repairs accounted for about 48% (RM 17,130) of the overall physical loss. In a similar vein, automobile repair expenditures came to almost 11% (RM 3900). As a result, the flood limits their mobility in addition to stopping their commercial operations. The Environment Agency's flood damage study showed that business property damages can be far greater than those of residential property [45].

Limitations and future research direction

There are some shortcomings in this study, even though the mapping of Shah Alam’s flood-prone areas is quite precise and useful. Consequently, further research should focus on the following limitations to give a more complete picture of the flooding in the region. In making decisions about flood control, this will assist Shah Alam. Rainfall was not considered in this examination as one of the flood conditioning issues because data on rainfall are only available from two locations. Not enough information on stream discharge rates to analyse flow rates and carry out hydrological studies. The lack of properly kept unpublished reports, old newspapers, and/or published peer-reviewed articles made it challenging to create an exhaustive inventory map of flood incidents. The field trips' sole objective was to identify flooded regions in real-time.

Conclusion

Effective and locally appropriate activities are needed to adapt to the changing global environment. Climate change adaptation strategies shouldn't involve isolated, segregated, or exclusive actions. Adaptive solutions must be incorporated into the current approaches to general catastrophe preparedness. When integrated into the total disaster preparedness measures, adaptation to climate change is instantly mainstreamed into the development plan in a local development planning process. The use of three MCDM models for identifying flood danger zones within the research region has been examined in this work. The AHP approach is the most widely used MCDM for mapping flood threats. For example, the proposed methodology can be applied to basin areas with different climate regimes and characteristics; a case study involving a different set or quantity of relevant FCFs; flood hazard maps derived from different dataset resolutions can be compared; and flood hazard maps derived from MCDM models other than the one used in this study can be compared to determine the best approach. Shah Alam is mostly a flood-affected area, and this study’s findings can help to identify the potential zones of flood in Shah Alam to protect the environment, pollution, and economic losses. To implement efficient flood management procedures, it can be concluded that the suggested models are a workable approach for locating flood hazards in this research region and other areas.

Acknowledgements

This study and publication are supported by the Ministry of Higher Education (MOHE) Malaysia, Fundamental Research Grant Scheme (FRGS)–FRGS/1/2022/SSI03/UKM/03/1.

Author contributions

Conceptualization: [Adam Narashman Leeonis], [Minhaz Farid Ahmed], [Mazlin Bin Mokhtar], [Rd Puteri Khairani Khirotdin]; Methodology: [Minhaz Farid Ahmed], [Rd Puteri Khairani Khirotdin], [Bijay Halder]; Formal analysis and investigation: [Adam Narashman Leeonis], [Minhaz Farid Ahmed], [Bijay Halder]; Writing—original draft preparation: [Adam Narashman Leeonis], [Minhaz Farid Ahmed], [Bijay Halder], [Chen Kim Lim], [Liew Juneng]; Writing—review and editing: [Adam Narashman Leeonis], [Minhaz Farid Ahmed], [Bijay Halder], [Chen Kim Lim], [Liew Juneng], [Mazlin Bin Mokhtar], [Rd Puteri Khairani Khirotdin]; Funding acquisition: [Minhaz Farid Ahmed]; Resources: [Minhaz Farid Ahmed], [Bijay Halder]; Supervision: [Minhaz Farid Ahmed], [Mazlin Bin Mokhtar].

Funding

This study and publication are supported by the Ministry of Higher Education (MOHE) Malaysia, Fundamental Research Grant Scheme (FRGS)–FRGS/1/2022/SSI03/UKM/03/1.

Data availability

“Data is provided within the manuscript”.

Code availability

Not applicable.

Declarations

Competing interests

The authors declare no competing interests.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Yaseen, ZM et al. Hourly river flow forecasting: application of emotional neural network versus multiple machine learning paradigms. Water Resour Manag; 2020; 34, 3 pp. 1075-1091. [DOI: https://dx.doi.org/10.1007/s11269-020-02484-w]

2. Sharir K, Rodeano R, Mariappan S. Flood susceptibility analysis (FSA) using analytical hierarchy process (AHP) model at The Kg. Kolopis area, Penampang, Sabah, Malaysia. In: Journal of Physics: Conference Series, IOP Publishing, 2019, p. 12065.

3. Tam, TH; Abd Rahman, MZ; Chan, NW; Woodhouse, I. Flood Risk Assessment for Kota Tinggi, Johor, Malaysia. Geospatial technology for water resource applications; 2016; Boca Raton, CRC Press: pp. 81-93.

4. Rehman, S et al. Analyzing vulnerability of communities to flood using composite vulnerability index: evidence from Bhagirathi Sub-basin, India. Nat Hazards; 2023; 119, 3 pp. 1341-1377. [DOI: https://dx.doi.org/10.1007/s11069-023-06170-z]

5. Agarwal, PK. Climate change: implications for Indian agriculture. Jalvigyan Sameeksha; 2007; 22, pp. 37-46.

6. Fotovatikhah, F; Herrera, M; Shamshirband, S; Chau, K; Faizollahzadeh Ardabili, S; Piran, MJ. Survey of computational intelligence as basis to big flood management: challenges, research directions and future work. Eng Appl Comput Fluid Mech; 2018; 12, 1 pp. 411-437. [DOI: https://dx.doi.org/10.1080/19942060.2018.1448896]

7. Rehman, S; Hasan, MSU; Rai, AK; Rahaman, MH; Avtar, R; Sajjad, H. Integrated approach for spatial flood susceptibility assessment in Bhagirathi sub-basin, India using entropy information theory and geospatial technology. Risk Anal; 2022; 42, 12 pp. 2765-2780. [DOI: https://dx.doi.org/10.1111/risa.13887]

8. Das Aditya, B. Flood risk reduction of Rupnarayana River, towards disaster management? A case study at Bandar of Ghatal Block in Gangetic Delta. J Geogr Nat Disasters; 2015; [DOI: https://dx.doi.org/10.4172/2167-0587.1000135]

9. Safiah Yusmah, MY et al. Understanding urban flood vulnerability and resilience: a case study of Kuantan, Pahang, Malaysia. Nat Hazards; 2020; 101, pp. 551-571. [DOI: https://dx.doi.org/10.1007/s11069-020-03885-1]

10. Tomar, P et al. GIS-based urban flood risk assessment and management—A case study of Delhi National capital territory (NCT), India. Sustainability; 2021; 13, 22 12850. [DOI: https://dx.doi.org/10.3390/su132212850]

11. Xiang, Z; Yan, J; Demir, I. A rainfall-Runoff Model With LSTM-based sequence-to-sequence learning. Water Resour Res; 2020; [DOI: https://dx.doi.org/10.1029/2019WR025326]

12. Kratzert, F; Klotz, D; Shalev, G; Klambauer, G; Hochreiter, S; Nearing, G. Towards learning universal, regional, and local hydrological behaviors via machine learning applied to large-sample datasets 2019; 10.5194/hess-23-5089-2019Hydrol Earth Syst Sci;

13. Elsheikh, RFA; Ouerghi, S; Elhag, AR. Flood risk map based on GIS, and multi criteria techniques (case study Terengganu Malaysia). J Geogr Inf Syst; 2015; 7, 04 348.

14. Khailani, DK; Perera, R. Mainstreaming disaster resilience attributes in local development plans for the adaptation to climate change induced flooding: a study based on the local plan of Shah Alam City, Malaysia. Land Use Policy; 2013; 30, 1 pp. 615-627. [DOI: https://dx.doi.org/10.1016/j.landusepol.2012.05.003]

15. Bui, DT et al. Novel hybrid evolutionary algorithms for spatial prediction of floods. Sci Rep; 2018; 8, 1 15364.[COI: 1:CAS:528:DC%2BC1MXhtFWmtbw%3D] [DOI: https://dx.doi.org/10.1038/s41598-018-33755-7]

16. Rahman, MR; Thakur, PK. Detecting, mapping and analysing of flood water propagation using synthetic aperture radar (SAR) satellite data and GIS: a case study from the Kendrapara District of Orissa State of India. Egypt J Remote Sens Sp Sci; 2018; 21, pp. S37-S41. [DOI: https://dx.doi.org/10.1016/j.ejrs.2017.10.002]

17. Saaty, TL. A scaling method for priorities in hierarchical structures. J Math Psychol; 1977; 15, 3 pp. 234-281. [DOI: https://dx.doi.org/10.1016/0022-2496(77)90033-5]

18. Ali, R; Sajjad, H; Masroor, M; Saha, TK; Roshani,; Rahaman, MH. Morphometric parameters based prioritization of watersheds for soil erosion risk in Upper Jhelum Sub-catchment, India. Environ Monit Assess; 2024; 196, 1 82. [DOI: https://dx.doi.org/10.1007/s10661-023-12226-1]

19. Saaty, TL. Decision making with the analytic hierarchy process. Int J Serv Sci; 2008; 1, 1 83. [DOI: https://dx.doi.org/10.1504/ijssci.2008.017590]

20. Ali, R; Sajjad, H; Saha, TK; Roshani,; Masroor, M; Rahaman, MH. Effectiveness of machine learning ensemble models in assessing groundwater potential in Lidder watershed, India. Acta Geophys; 2024; 72, 4 pp. 2843-2856. [DOI: https://dx.doi.org/10.1007/s11600-023-01237-8]

21. Aydin, HE; Iban, MC. Predicting and analyzing flood susceptibility using boosting-based ensemble machine learning algorithms with SHapley Additive exPlanations. Nat Hazards; 2022; [DOI: https://dx.doi.org/10.1007/s11069-022-05793-y]

22. Bhuyan, N et al. Assessing landscape ecological vulnerability to riverbank erosion in the Middle Brahmaputra floodplains of Assam, India using machine learning algorithms. CATENA; 2024; 234, 107581. [DOI: https://dx.doi.org/10.1016/j.catena.2023.107581]

23. Saaty, TL. The analytic network process—dependence and feedback in decision-making. Bus Appl Comput Intell; 2006; [DOI: https://dx.doi.org/10.4018/978-1-59140-702-7.ch018]

24. Chakrabortty, R et al. Flash-flood hazard susceptibility mapping in Kangsabati River Basin, India. Geocarto Int; 2021; 37, 23 pp. 6713-6735. [DOI: https://dx.doi.org/10.1080/10106049.2021.1953618]

25. Mitra, R; Das, J. A comparative assessment of flood susceptibility modelling of GIS-based TOPSIS, VIKOR, and EDAS techniques in the Sub-Himalayan foothills region of Eastern India. Environ Sci Pollut Res; 2023; 30, 6 pp. 16036-16067. [DOI: https://dx.doi.org/10.1007/s11356-022-23168-5]

26. Houghton-Carr, HA; Print, CR; Fry, MJ; Gadain, H; Muchiri, P. An assessment of the surface water resources of the Juba-Shabelle basin in southern Somalia. Hydrol Sci J; 2011; 56, 5 pp. 759-774. [DOI: https://dx.doi.org/10.1080/02626667.2011.585470]

27. Patrikaki, O; Kazakis, N; Kougias, I; Patsialis, T; Theodossiou, N; Voudouris, K. Assessing flood hazard at river basin scale with an index-based approach: the case of Mouriki, Greece. Geosciences; 2018; 8, 2 50. [DOI: https://dx.doi.org/10.3390/geosciences8020050]

28. Pawar, U; Suppawimut, W; Muttil, N; Rathnayake, U. A GIS-based comparative analysis of frequency ratio and statistical index models for flood susceptibility mapping in the upper Krishna Basin, India. Water; 2022; 14, 22 3771. [DOI: https://dx.doi.org/10.3390/w14223771]

29. Ghosh, A; Kar, SK. Application of analytical hierarchy process (AHP) for flood risk assessment: a case study in Malda district of West Bengal, India. Nat Hazards; 2018; 94, 1 pp. 349-368. [DOI: https://dx.doi.org/10.1007/s11069-018-3392-y]

30. Saaty, TL. How to make a decision: the analytic hierarchy process. Eur J Oper Res; 1990; 48, 1 pp. 9-26. [DOI: https://dx.doi.org/10.1016/0377-2217(90)90057-i]

31. Nachappa, TG; Piralilou, ST; Gholamnia, K; Ghorbanzadeh, O; Rahmati, O; Blaschke, T. Flood susceptibility mapping with machine learning, multi-criteria decision analysis and ensemble using Dempster Shafer Theory. J Hydrol; 2020; 590, 125275. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2020.125275]

32. Yalew, SG; van Griensven, A; Mul, ML; van der Zaag, P. Land suitability analysis for agriculture in the Abbay basin using remote sensing, GIS and AHP techniques. Model Earth Syst Environ; 2016; [DOI: https://dx.doi.org/10.1007/s40808-016-0167-x]

33. Getis, A; Ord, JK. The analysis of spatial association by use of distance statistics. Geogr Anal; 1992; 24, 3 pp. 189-206. [DOI: https://dx.doi.org/10.1111/j.1538-4632.1992.tb00261.x]

34. Yang, S; Taylor, D; Yang, D; He, M; Liu, X; Xu, J. A synthesis framework using machine learning and spatial bivariate analysis to identify drivers and hotspots of heavy metal pollution of agricultural soils. Environ Pollut; 2021; 287, 117611.[COI: 1:CAS:528:DC%2BB3MXhsVWhsbfJ] [DOI: https://dx.doi.org/10.1016/j.envpol.2021.117611]

35. Mudashiru, RB; Sabtu, N; Abdullah, R; Saleh, A; Abustan, I. A comparison of three multi-criteria decision-making models in mapping flood hazard areas of Northeast Penang, Malaysia. Nat Hazards; 2022; 112, 3 pp. 1903-1939. [DOI: https://dx.doi.org/10.1007/s11069-022-05250-w]

36. Chakraborty, S; Mukhopadhyay, S. Assessing flood risk using analytical hierarchy process (AHP) and geographical information system (GIS): application in Coochbehar district of West Bengal, India. Nat Hazards; 2019; 99, 1 pp. 247-274. [DOI: https://dx.doi.org/10.1007/s11069-019-03737-7]

37. FDTCP. (2012). Plan Malaysia Annual Report 2012. Kuala Lumpur, Malaysia: Federal Department of Town and Country Planning (FDTCP). Available: https://www.planmalaysia.gov.my/planmalaysia/resources/Ranacangan/Laporan%20Tahunan_Buletin/annual_report_2012.pdf

38. Bin Mokhtar, M; Ahmed, MF. Managing a road as a river to mitigate the impact of urban flash floods. J Flood Risk Manag; 2022; 15, 4 pp. 1-2.

39. Rosmadi, HS; Ahmed, MF; Bin Mokhtar, M; Lim, CK. Reviewing challenges of flood risk management in Malaysia. Water; 2023; 15, 13 2390. [DOI: https://dx.doi.org/10.3390/w15132390]

40. Bari, MA; Alam, L; Alam, MM; Rahman, LF; Pereira, JJ. Estimation of losses and damages caused by flash floods in the commercial area of Kajang, Selangor, Malaysia. Arab J Geosci; 2021; 14, pp. 1-9. [DOI: https://dx.doi.org/10.1007/s12517-021-06503-x]

41. Krishnan, G. (2008). Worst flood in Kajang. The Star Online. Star Media Group Berhad. Available: https://www.thestar.com.my/news/community/2008/10/17/worst-flood-inkajang

42. Lawal DU, Matori A-N, Yusuf KW, Hashim AM, Balogun A-L. Analysis of the flood extent extraction model and the natural flood influencing factors: A GIS-based and remote sensing analysis. In: IOP conference series: earth and environmental science, IOP Publishing, 2014, p. 12059.

43. Asmara TAT, Ludin ANM. Mapping perception of community preparedness towards flood in Muar river, Johor Malaysia. In IOP Conference Series: Earth and Environmental Science, IOP Publishing, 2014, p. 12126.

44. Noorazuan MH, Rainis R, Juahir H, Zain SM, Jaafar N. GIS application in evaluating land use-land cover change and its impact on hydrological regime in Langat River basin, Malaysia. In: 2nd annual Asian conference of map Asia, Citeseer, 2003, pp. 14–15.

45. Environment Agency. (2018). Estimating the economic costs of the 2015 to 2016 winter floods. Bristol, UK: Environment Agency. Available: https://assets.publishing.service.gov.uk/media/5a755ce8ed915d7314959615/Estimating_the_economic_costs_of_the_winter_floods_2015_to_2016.pdf