# Comprehensive Assessment of Flood Hazard, Vulnerability, and Flood Risk at the Household Level in a Municipality Area: A Case Study of Nan Province, Thailand

^{1}

^{2}

^{*}

## Abstract

**:**

## 1. Introduction

- (a)
- To simulate floods in the Upper Nan River and its floodplain in the municipal area of Nan Province;
- (b)
- To develop a comprehensive and systematic methodology to determine flood hazard, flood vulnerability, and flood risk in a municipal area at the household level;
- (c)
- To apply the developed methodology to assess flood hazard, flood vulnerability, and flood risk at the household level in the Nan Municipality area considering floods of various return periods;
- (d)
- To analyze and discuss the consequences of flood hazard, vulnerability, and flood risk on local residents and the physical environment.

## 2. Study Area

^{2}[26]. Its main drainage way is the Nan River. It has an annual average temperature of 26.3 °C, relative humidity of 75.9%, wind speed of 0.9 knots, evaporation of 1457.4 mm, and rainfall of 1371 mm.

^{2}[27]. Another river gauging station upstream of station N1 is the Tha Wang Pha station (N64). The basin study area is subdivided into five sub-basins, namely, Upper Nan, Nam Yao (W), Nam Yao (E), Nam Samun, and Nam Nan part-2 as shown in Figure 2 [28].

^{2}on the right bank of the Nan River as shown in Figure 3. The municipal area is an urban area that has 31 villages that are mainly residential. The Nan River flows through the municipal area for a distance of 7 km. In this river reach, overbank flow occurs frequently during high flood periods. The municipal area has flooded often in the past. Large flood inundations occurred in 2006 [29] and 2011 due to the fact of heavy tropical storms [30]. Considerable flood damages and casualties were reported in both years.

## 3. Research Structure and Methodology

#### 3.1. Research Structure

#### 3.2. Methodology

#### 3.2.1. Determination of Flood Hazard

#### 3.2.2. Determination of Total Flood Damage Vulnerability

_{pop}, and the weighted flood vulnerability index of households, FVI

_{hh}. The total vulnerability index (FVI) of each village is calculated by the following equations:

_{pop}and w

_{hh}are determined using AHP. The level of total vulnerability FVI is specified as very low when FVI ≤ 1, low when 1 < FVI ≤ 2, medium when 2 < FVI ≤ 3, high when 3 < FVI ≤ 4, and very high when 4 < FVI ≤ 5.

_{pop}, population vulnerability, VI

_{pop}, is represented by population density (person/km

^{2}) and is classified into five ranges in ascending order as given in Table 2. These five ranges are represented by the integer numbers 1, 2, 3, 4, and 5 for the index FVI

_{pop}: very low, low, medium, high, and very high respectively. The household vulnerability index, FVI

_{hh}, is determined by the value of the household vulnerability. VI

_{hh}and its ranges according to Table 2. The VI

_{hh}of the household samples is expressed as a function of three major contributing factors, namely: sensitivity F1, adaptive capacity F2, and exposure F3 according to the following equation:

_{1}, w

_{2}, and w

_{3}are the weights of F1, F2, and F3, respectively. The major contributing factors F1, F2, and F3 are defined in [24,25]. Equation (5) was developed in this study based on the original equation in [34,35,36] by including the weights w

_{1}, w

_{2}, and w

_{3}in Equation (5). The weighting factors w

_{1}, w

_{2}, and w

_{3}were introduced to normalize the relative importance of F1, F2, and F3. The values of w

_{1}, w

_{2}, and w

_{3}were determined by the AHP using data from collected samples of questionnaires and field surveys. Equation (6) requires that the sum of the weighting factors w

_{1}, w

_{2}and w

_{3}is equal to one.

_{1}–

_{7}in which C

_{1}= family size, C

_{2}= gender, C

_{3}= health, C

_{4}= land use type, C

_{5}= household damage, C

_{6}= public property damages, and C

_{7}= ownership of household. The sensitivity F1 is computed as:

_{i}is the ith contributing component of the sensitivity factor F1, and n = 7 is the total number of the component C

_{i}of F1.

_{i}of F1 is computed as:

_{i}of C

_{i}is determined by AHP based on the collected samples from questionnaires and field surveys; m is the number of classes of the collected samples for the component C

_{i}, arranged from the highest to the lowest significance of vulnerability; Q

_{j}is the number of samples of class j as a percentage of the total collected samples of all classes; K

_{j}is the assigned impact score of the class j between 0% and 100%. A score of 100% is assigned to the class that has the highest impact on vulnerability. More details regarding the calculation are given in the next section on the computational procedure.

_{1}for F1 is calculated as follows:

_{j}is the number of family samples in the jth class with the assigned impact score K

_{j}, and θ

_{1}is the weight of C

_{1}to be determined by AHP.

_{hh}. The computed household vulnerability, VI

_{hh}, of each village was fitted into the five classified ranges: very low, low, medium, high, and very high as shown in Table 2. Then, the value of FVI

_{hh}corresponding to VI

_{hh}was represented by an integer number from 1 to 5 according to Table 2.

_{pop}, and of household FVI

_{hh}were substituted into Equation (3) to compute the total vulnerability, VI.

#### 3.2.3. Determination of Flood Risk

_{hh}, and the total vulnerability, FVI. The computed values of the FRI were classified into five ranges corresponding to very low for 1 < FRI ≤ 5, low for 5 < FRI ≤10, medium for 10 < FRI ≤ 15, high for 15 < FRI ≤ 20, and very high for 20 < FRI ≤ 25.

#### 3.3. Data Collection

#### 3.3.1. Hydrological Data

#### 3.3.2. Vulnerability Data

#### 3.4. Computational Procedure

#### 3.4.1. Computation of Flood Hazard

- (a)
- Rainfall–runoff computation: The HEC-HMS rainfall–runoff model [38] was applied to compute the runoff hydrograph using hourly rainfall input at four stations in the Upper Nan Basin. The hourly rainfalls at the four stations were averaged over the basin area using the Thiessen polygon method. The river basin was divided into seven sub-basins in which the hourly average rainfalls were used in each sub-basin. The computed runoff was used as the upstream boundary condition of the HEC-RAS flood routing model [39]. The HEC-HMS model requires a digital elevation model (DEM), soil and land-use maps, soil characteristics, and input rainfall hyetographs. The HEC–Geo HMS model, which is an extension of HEC-HMS, prepares raster layers of delineated sub-basins and river network systems for exporting to HEC-HMS as base maps. By inputting rainfall data, land cover, and soil maps to HEC-HMS, the model computes daily runoff hydrographs for each sub-basin. The HEC-HMS model was calibrated and verified against the observed daily discharges at station N64 at Tha Wang Pha and at station N1 at Muang Nan. The calibration period was during the wet period from June to December 2006–2011, and the verification period was from June to December 2012–2017. In the model calibration, the model parameters, such as initial and maximum storages of canopy, SCS curve number, time of concentration, and lag of unit hydrograph, were assumed and adjusted by trial and error to obtain a satisfactory agreement between the observed and computed discharge hydrographs;
- (b)
- Flood routing computation: The 1D and 2D HEC-RAS flood routing model [39] for the Upper Nan River and its floodplain were used to route the runoff from the upstream station N1 along the Nan River to the downstream end station, which is 7 km downstream of station N1. The river passes through the municipal area, which is in the river floodplain. The geometrical inputs to HEC-RAS were the measured river cross-sections every 1.2 km and the floodplain topography from the digital elevation model with a 30 by 30 m resolution with a 1 m contour interval. The river cross-sections from the field measurements and the floodplain topography from the DEM were merged using HEC-GEO RAS, an extension of HEC-RAS, to obtain the complete river and floodplain cross-sections [40]. This geometry data was input into HEC-RAS for flood flow simulation. In the HEC-RAS model, the 1D flow routing procedure was used to compute the 1D flow in the Nan River, and the 2D flow routing was used to compute the 2D flow in the floodplain;
- (c)
- The hydrological inputs to HEC-RAS were the observed daily upstream discharge at station N1 as the upstream boundary condition. At the downstream end of the model, there was no river gauging station; thus, a depth–discharge relationship according to the Manning equation was used. The model was calibrated and verified by trial-and-error adjustment based on the values of the Manning roughness coefficient n;
- (d)
- Calculation of flood hazard index: The FHI was computed for each grid of 30 × 30 m in the municipality’s floodplain area using Equation (1). The flood duration index (FHI
_{T}), the depth index (FHI_{D}), and the velocity index (FHI_{V}) were determined by using the results of the HEC-RAS model and the classification in Table 3. As shown in Figure 4, for Blocks A.1 to A.4, the computed FHI_{D}, FHI_{V}, and FHI_{T}were substituted into Equation (1) to compute the FHI. The hazard weighting factors α for flood duration, β for flood depth, and µ for flood velocity in Equation (1) were determined using AHP [24,25].

#### 3.4.2. Computation of Total Vulnerability

^{2}) of each village from census data. For the household vulnerability, VI

_{hh}, each major contributing factor of VI

_{hh}was composed of a number of components in which each component was further subdivided into classes with assigned impact scores. For example, the major contributing factor F1, or sensitivity, had seven components (C

_{1}–

_{7}), namely: family size, gender, health, land-use type, household damage, public damage, and household ownership as shown in Table 3. An example of calculating the component C

_{1}or family size of F1 in Table 3 for Phumin-Thali village is given here. The family size was subdivided into three classes: class 1 for households with more than five family members; class 2 for between three and five family members; class 3 for fewer than three family members. The impact scores of 100%, 67%, and 33% were assigned to each family class according to the information from local residents on vulnerability to their families. A score of 100% was assigned to class 1, because it is most sensitive to vulnerability. The number of collected household samples in Phumin-Thali village was six.

_{1}of the major contributing factor F1 was classified into three classes (m = 3). In class 1 (j = 1), the number of the families was equal to one. Hence, Q

_{1}was equal to 16.67% of the total collected samples of six, and the impact score K

_{1}was assigned to be 100%. In class 2 (j = 2), the number of families was equal to two or Q

_{2}= 33.33% and assigned K

_{2}= 67%. In class 3 (j = 3), the number of families was three or Q

_{3}= 50% and K

_{3}= 33%. The weighting factor θ

_{1}of the sensitivity factor F1 was determined by AHP and equal to 0.386. This is shown at the beginning of Table 3. Knowing Q

_{1}, Q

_{2}, Q

_{3}, K

_{1}, K

_{2}, K

_{3}, and θ

_{1}, the value of the component C

_{1}of F1 is equal to θ

_{1}(Q

_{1}K

_{1}+ Q

_{2}K

_{2}+ Q

_{3}K

_{3})/100 = 21.41%.

_{1}of the factor F1 as shown in Table 3. In this way, the components C

_{1}–

_{7}were calculated using Equation (8) and summed to obtain the major contributing factor F1 according to Equation (7).

#### 3.4.3. Computation of Flood Risk

## 4. Results

#### 4.1. Calibration and Verification of the HEC-HMS Rainfall–Runoff Model and HEC-RAS Flood Routing Model

^{2}), Nash–Sutcliffe efficiency (NSE), percent bias (PBIAS), volume ratio (Vr), and normalized root mean square error (NRMSE). The results of the model’s calibration at stations N64 and N1 were found to be satisfactory with coefficients of determination of 0.72 and 0.76; Nash–Sutcliffe efficiencies of 0.67 and 0.64; percent biases of 15.78 and 15.11; volume ratios of 0.95 and 0.89; normalized root mean square errors of 12.29% and 14.36%, respectively. The statistics showed acceptable agreement in both the model’s calibration and verification.

^{−1}.

#### 4.2. Flood Hazard

^{2}, the total flooded area for the 100 year return period was found to be 1,614,545 km

^{2}, or 21.24% of the study area. The remaining part of the study area had no flood inundation. According to the ranges of classification, the hazard index of the flooded area of 1,614,545 km

^{2}was classified into a very high hazard area of 307,638 km

^{2}, or 19.05% of the study area; a high hazard of 25,622 km

^{2}, or 1.59%; a medium hazard of 68,432 km

^{2}, or 4.32%; a low hazard of 1,014,760 km

^{2}, or 62.85%; a very low hazard of 198,094 km

^{2}, or 12.26%. The high and very hazardous areas were in Phumin-Thali village in the south of the study area, and another major flooding location was in the area of Suriyapong Camp in the north of the study area.

#### 4.3. Total Flood Vulnerability

_{pop}, and the computed household vulnerability index, FVI

_{hh}, were substituted into Equation (3). The weights w

_{pop}for population vulnerability and w

_{hh}for household vulnerability were determined using AHP based on the data collected from the Nan Municipality and from the questionnaire survey. The weights w

_{pop}and w

_{hh}were found to be 0.33 and 0.67, respectively. This shows that the household vulnerability, FVI

_{hh}, has much more influence on the total vulnerability, FVI, than the population vulnerability, FVI

_{pop}.

#### 4.4. Flood Risk

## 5. Discussions

- The performance of the HEC-HMS model was evaluated using the following statistical parameters, namely: R
^{2}, NSE, PBIAS, Vr, and NRMSE. The results of the model’s calibration at stations N64 and N1 were found to be satisfactory as shown in the results, and. the performance statistics showed acceptable agreement in both the model’s calibration and verification; - The comparison of the computed and observed inundation depths of high-water marks in the floodplains at Phuang-Payom, Phumin-Thali, and Mueang Len villages were found to be satisfactory with percentage errors of 6.7, 6.3, and 9.1, respectively. This assures the accuracy of the model’s simulation in the floodplain. Importantly, the results show the reliability of the HEC-RAS flood simulation model and the data used in the calculation;
- The flood hazard was calculated by using Equations (1) and (2) in which the weights α, β, and μ representing flood duration, depth, and velocity, respectively, were systematically determined by AHP. In the other previous studies [17,18,20], the weights α, β, and μ for flood duration, depth, and velocity, respectively, in Equation (1) were specified by the researchers according to their experiences without using AHP. These weights could be subjective or questionable. By using the AHP, the weights α, β, and μ of 0.63, 0.26, and 0.11, respectively, were obtained. In AHP, the relative importance of a flood duration, depth, and velocity of 1, 3, and 5 was assigned according to the results of field and questionnaire surveys. A sensitivity analysis was conducted to determine the effects of change on the relative importance of flood duration, depth and, velocity from the original values of 1, 3, and 5 to the new values of 1, 5, and 9, respectively. By using AHP, the new weights α, β, and μ were found to be 0.72, 0.21 and 0.06, respectively. In percent, the changes were +14.3% for weight α, −19.2% for β, and −45.4% for μ. The changes in weights α and β were considered to be small and acceptable. The change in weight μ of −45.4% was negative and moderate. Based on the field and questionnaire surveys in this study, it was revealed that the damaging effect of flow velocity in the municipal area was much smaller compared to flood duration and depth; therefore, the change in weight μ of −45.4% was considered non-significant. Hence, weights α, β, and μ of 0.63, 0.26, and 0.11, respectively, were considered reasonable;
- As shown in Figure 8 and Figure 9, the flood hazard increased both spatially and in magnitude with the increase in the flood return periods. The 10 year flood hazard along the right bank of the Nan River was smaller than the 50 and 100 year floods, and it was much smaller compared to the 500 year flood. Significant flood hazard occurs in the Phumin-Thali and Phuang-Payom villages in the southern part of the municipality. The locations of the observed high-water marks are shown by the small red circles in Figure 10. It can be seen that the center of the municipal area has higher ground elevation than the surrounding area and, hence, it has less of a flood hazard. On the other hand, in the Mueang Len area in the northeastern part of the municipality, the hazard is significant when the flood magnitude is greater than the 100 year return period;
- For total flood vulnerability, the weights W
_{pop}and W_{hh}in Equation (3) were found to be 0.33 and 0.67, respectively. This shows that FVI_{hh}had much more influence on FVI than FVI_{pop}. For household vulnerability, FVI_{hh}, the relative importance of the major contributing factors, namely, sensitivity F1, adaptation F_{2}, and exposure F_{3}were set to be 1, 3, and 5, respectively. By using AHP, the weights of w_{1}of sensitivity F1, w_{2}of adaptation F2, and w_{3}of exposure F3 were found to be 0.63, 0.26, and 0.11, respectively. The same sensitivity analysis was conducted for the case of flood hazard, and it was found that the values of 0.63, 0.26, and 0.11 for weights w_{1}, w_{2}, and w_{3}, respectively, were the most reasonable ones. In previous studies [21,23], the weights w_{1}, w_{2}, and w_{3}were not determined by AHP but were assumed to be equal to one. Such an assumption could be incorrect, as the values of the sensitivity F1, the adaptation F2, and the exposure F3 were calculated on different bases, and they were not normalized. In this study, each major contributing factor was considered to have various contributing components Ci as shown in Table 3. The weight θi of each component Ci was determined by AHP based on the collected samples from the questionnaire surveys as given in Table 3; - The distribution of the total vulnerability index, FVI, in the municipal area is shown in Figure 10. Depending on the sensitivity, adaptive capacity, and exposure of the households and population, the very high and highly vulnerable areas were found along the right bank of the Nan River, from upstream to downstream. These areas included Phumin-Thali, Phuang-Payom, and the Mueang Len villages. The medium vulnerable areas are in the central and western parts. Only two villages in the western rim of the municipal area, namely, Pha Mai and Don Swan, have very low vulnerability, because they have very low population densities and are located on higher ground elevations, far away from the river;
- Flood risks depend on flood hazard probabilities and vulnerabilities. Therefore, flood risks also change with flood probabilities or flood return periods. The study’s results show that when the flood hazard changes, the flood risk also changes correspondingly;
- To mitigate flood problems in the municipal area, various flood control or mitigation measures can be proposed such as dredging of the Nan River channel and its tributaries, raising crest elevations of river flood control levees, or construction of flood bypass channel around the Nan municipal area. The effectiveness of each measure in reducing the flood hazard can be evaluated by using the hydrological model (HEC-HMS) and the hydrodynamic flood routing model (HEC-RAS). Hence, the changes in flood risk in Nan Municipality can be determined. More details can be found in [21].

## 6. Conclusions

- The novelty of this study is the development of an advanced comprehensive and systematic methodology in determining flood hazard, total flood vulnerability, and flood risk at the household level in a municipal area. This is an important improvement over previous studies in which the flood and vulnerability parameters were not all considered. Moreover, these parameters were not systematically determined;
- The methodology was applied to a case study of a municipal area of 7.6 km
^{2}in Nan Province, Northern Thailand. The study area was located in the floodplain on the right bank of the Nan River; - The HEC-HMS hydrological model and the HEC-RAS flood flow simulation model were applied to predict flood depths, velocities, and durations for the return periods of 10, 50, 100, and 500 years;
- The model computed results showed that significant flood hazards occur in Phumin-Thali and Phuang-Payom villages in the southern part of the municipal area and in Mueang Len village in the northeastern part of the area. These villages have low ground elevations, and they are near to Nan River. The central part of the municipal area had less of a flood hazard, as it has a higher ground elevation, and it is far from the river. The computed results were found to be consistent with the past flood situations during the field survey;
- The questionnaire survey in the municipal area reported that the flood duration had the most significant impact on households, while the flood depth and velocity had lesser impacts;
- In-depth analysis of the total vulnerability in the municipal area showed that the vulnerability of the population constituted one-third of the total vulnerability, while the household vulnerability constituted the remaining two-thirds;
- From the computed flood risks, flood risk maps were constructed for various return periods. The maps show that Phumin-Thali and Phuang-Payom villages, located near the right bank of the Nan River, are under very high risk, and more than half of the villages will be inundated and prone to high flood damages. This is consistent with past flood situations;
- The flood risk in the municipal area increased by approximately four times for the increase in the return period from 10 to 500 years;
- Overall, the methodology developed in this study yielded realistic results, and it should be applied to other study areas.

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Acknowledgments

## Conflicts of Interest

## Appendix A

Judgment Index | Flood Duration | Water Depth | Flow Velocity |
---|---|---|---|

Flood Duration | 1 | 3 | 5 |

Water Depth | $\frac{1}{3}$ | 1 | 3 |

Flow Velocity | $\frac{1}{5}$ | $\frac{1}{3}$ | 1 |

Sum | $\frac{23}{15}$ | $\frac{13}{3}$ | 9 |

Judgment Index | Flood Duration | Water Depth | Flow Velocity | Normalized Weights (WI) |
---|---|---|---|---|

Flood Duration | $\frac{15}{23}$ | $\frac{9}{13}$ | $\frac{5}{9}$ | 0.63 |

Water Depth | $\frac{5}{23}$ | $\frac{3}{13}$ | $\frac{3}{9}$ | 0.26 |

Flow Velocity | $\frac{3}{23}$ | $\frac{1}{13}$ | $\frac{1}{9}$ | 0.11 |

Sum | 1.00 | 1.00 | 1.00 | 1.00 |

_{max}), consistency ratios (CR), and the consistency index (CI).

_{max}, the CR, and the CI were computed using the following equations:

_{max}is the principal eigenvalue, n is the number of rows or columns of the square matrix, a

_{ij}is the element of judgment matrix, W

_{i}is the normalized weight of row i in Table A2, CR is the consistency ratio, CI is the consistency index, and RI is the random inconsistency index according to [24,25].

_{max}) computed using Equation (A1) was equal to (23/15 × 0.63 + 13/3 × 0.26 + 9 × 0.11) = 3.083. This was approximately equal to three, which is the number of flood hazard parameters. The consistency index and consistency ratio should be approximately zero for perfect weights. The CI was computed by Equation (A2) as:

## References

- World Health Organization. Home/Health Topics/Floods. 2021. Available online: https://www.who.int/health-topics/floods (accessed on 7 November 2021).
- The Organization for Economic Co-operation and Development. Executive summary. In Financial Management of Flood Risk; OECD Publishing: Paris, France, 2016. [Google Scholar] [CrossRef]
- Ritchie, H.; Roser, M. Natural Disasters. Published Online at OurWorldInData.org. 2014. Available online: https://ourworldindata.org/natural-disasters (accessed on 7 November 2021).
- Statista. 2021. Available online: https://www.statista.com/statistics/267746/number-of-deaths-globally-due-to-major-flooding/ (accessed on 7 November 2021).
- Arnell, N.W.; Gosling, S.N. The impacts of climate change on river flood risk at the global scale. Clim. Chang.
**2016**, 134, 387–401. [Google Scholar] [CrossRef] [Green Version] - United Nations Office for Disaster Risk Reduction, UNISDR. Flood Hazard and Risk Assessment; United Nations Office for Disaster Risk Reduction: Geneva, Switzerland, 2017; Available online: https://www.preventionweb.net/files/52828_04floodhazardandriskassessment.pdf (accessed on 15 May 2020).
- Quesada-Román, A.; Ballesteros-Cánovas, J.A.; Granados-Bolaños, S.; Birkel, C.; Stoffel, M. Dendrogeomorphic reconstruction of floods in a dynamic tropical river. Geomorphology
**2020**, 359, 107133. [Google Scholar] [CrossRef] - Quesada-Román, A.; Villalobos-Chacón, A. Flash flood impacts of Hurricane Otto and hydrometeorological risk mapping in Costa Rica. Geogr. Tidsskr.-Dan. J. Geogr.
**2020**, 120, 142–155. [Google Scholar] [CrossRef] - Tingsanchali, T. Urban flood disaster management. Procedia Eng.
**2012**, 32, 25–37. [Google Scholar] [CrossRef] [Green Version] - Quesada-Román, A.; Ballesteros-Cánovas, J.A.; Granados-Bolaños, S.; Birkel, C.; Stoffel, M. Improving regional flood risk assessment using flood frequency and dendrogeomorphic analyses in mountain catchments impacted by tropical cyclones. Geomorphology
**2022**, 396, 108000. [Google Scholar] [CrossRef] - Kvočka, D.; Falconer, R.A.; Bray, M. Flood hazard assessment for extreme flood events. Nat. Hazards
**2016**, 84, 1569–1599. [Google Scholar] [CrossRef] [Green Version] - Quesada-Román, A.; Castro-Chacón, J.P.; Feoli-Boraschi, S. Geomorphology, land use, and environmental impacts in a densely populated urban catchment of Costa Rica. J. S. Am. Earth Sci.
**2021**, 112, 103560. [Google Scholar] [CrossRef] - García-Soriano, D.; Quesada-Román, A.; Zamorano-Orozco, J.J. Geomorphological hazards susceptibility in high-density urban areas: A case study of Mexico City. J. S. Am. Earth Sci.
**2020**, 102, 102667. [Google Scholar] [CrossRef] - Quesada-Román, A.; Villalobos-Portilla, E.; Campos-Durán, D. Hydrometeorological disasters in urban areas of Costa 13 Rica, Central America. Environ. Hazards
**2020**, 20, 264–278. [Google Scholar] [CrossRef] - Engle, N.L. Adaptive capacity and its assessment. Glob. Environ. Chang.
**2011**, 21, 647–656. [Google Scholar] [CrossRef] - Kron, W. Flood risk = hazard·values·vulnerability. Water Int.
**2005**, 30, 58–68. [Google Scholar] [CrossRef] - Tingsanchali, T.; Karim, F. Flood-hazard assessment and risk-based zoning of a tropical flood plain: Case study of the Yom River, Thailand. Hydrol. Sci. J.
**2010**, 55, 145–161. [Google Scholar] [CrossRef] - Keokhumcheng, Y.; Tingsanchali, T.; Clemente, R.S. Flood risk assessment in the region surrounding the Bangkok Suvarnabhumi Airport. Water Int.
**2012**, 37, 201–217. [Google Scholar] [CrossRef] - Vojtek, M.; Vojteková, J. Flood hazard and flood risk assessment at the local spatial scale: A case study. Geomatics. Nat. Hazards Risk
**2016**, 7, 1973–1992. [Google Scholar] [CrossRef] - Tingsanchali, T.; Keokhumcheng, Y. A method for evaluating flood hazard and flood risk of east Bangkok plain, Thailand. Proc. Inst. Civ. Eng.—Eng. Sustain.
**2019**, 172, 385–392. [Google Scholar] [CrossRef] - Promping, T. Integrated Flood Risk Assessment and Management with Flood Control Measures in Nan Municipality, Thailand. Master’s Thesis, Asian Institute of Technology, Khlong Nueng, Thailand, 2019; 137p. [Google Scholar]
- Department of Water Resources. Integrated Plans for Water Resources Management in the Nan River Basin; Department of Water Resources: Bangkok, Thailand, 2003. [Google Scholar]
- Samarasinghe, S.M.J.S.; Nandalal, H.K.; Welivitiya, D.P.; Fowze, J.S.M.; Hazarika, M.K.; Samarakoon, L. Application of Remote Sensing and GIS for Flood Risk Analysis: A Case Study at Kalu- Ganga River, Sri Lanka. In Proceedings of the International Archives of the Photogrammetry, Remote Sensing and Spatial Information Science, Kyoto, Japan, 9–12 August 2010; Volume XXXVIII. Part 8. [Google Scholar]
- Saaty, T.L. Risk-Its priority and probability: The Analytic Hierarchy Process. Risk Anal.
**1987**, 7, 159–172. [Google Scholar] [CrossRef] - Saaty, T.L. Fundamentals of Decision Making and Priority Theory with the Analytic Hierarchy Process AHP; RWS Publications: Pittsburg, PA, USA, 2000. [Google Scholar]
- Wangpimool, W.; Pongput, K.; Sukvibool, C.; Sombatpanit, S.; Gassman, P.W. The effect of reforestation on stream flow in Upper Nan river basin using Soil and Water Assessment Tool (SWAT) model. Int. Soil Water Conserv. Res.
**2013**, 1, 53–63. [Google Scholar] [CrossRef] [Green Version] - Department of Disaster Prevention and Mitigation (DDPM). A Development Network of Natural Disasters for Flash Floods and Landslides; Project Final Report; Institute of Science and Technology, Chiang Mai University: Chiang Mai, Thailand, 2010. [Google Scholar]
- Royal Irrigation Department (RID). Flood Report in the Nan Basin, Thailand [in Thai]; Royal Irrigation Department: Bangkok, Thailand, 2006. [Google Scholar]
- Poaponsakorn, N.; Meethom, P. Impact of the 2011 Floods, and Flood Management in Thailand; ERIA Discussion Paper Series, ERIA-DP-2013-34; Thailand Development Research Institute, Economic Research Institute for ASEAN and East Asia: Bangkok, Thailand, 2013; Available online: https://www.eria.org/ERIA-DP-2013-34.pdf (accessed on 18 March 2021).
- Geo-Informatics and Space Technology Development Agency (GISTDA). Radar Satellite Images and Flood Maps of the 2011 Flood; Geo-Informatics and Space Technology Development Agency: Bangkok, Thailand, 2011. [Google Scholar]
- Seejata, K.; Yodying, A.; Wongthadam, T.; Mahavik, N.; Tantanee, S. Assessment of flood hazard areas using Analytic Hierarchy Process over the Lower Yom Basin, Sukhothai Province. Procedia Eng.
**2018**, 212, 340–347. [Google Scholar] [CrossRef] - Ghosh, A.; Kar, S.K. Application of analytic hierarchy process (AHP) for flood risk assessment: A case study in Malda district of West Bengal, India. Nat. Hazards
**2018**, 94, 349–368. [Google Scholar] [CrossRef] - Alfa, M.I.; Ajibike, M.A.; Daffi, R.E. Application of Analytic Hierarchy Process and Geographic Information System Techniques in Flood Risk Assessment: A Case of Ofu River Catchment in Nigeria. J. Degrad. Min. Lands Manag.
**2018**, 5, 1363–1372. [Google Scholar] [CrossRef] [Green Version] - Intergovernmental Panel on Climate Change, IPCC. Climate Change 2007: Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report (Chapter 9); Cambridge University Press: Cambridge, UK, 2007. [Google Scholar]
- Intergovernmental Panel on Climate Change, IPCC. Climate Change 2007: The physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report (Chapter 11); Cambridge University Press: Cambridge, UK, 2007. [Google Scholar]
- Hahn, M.B.; Riederer, A.M.; Foster, S.O. The livelihood vulnerability index: A pragmatic approach to assessing risks from climate variability and change—A case study in Mozambique. Glob. Environ. Chang.
**2009**, 19, 74–88. [Google Scholar] [CrossRef] - Yamane, T. Statistics, An Introductory Analysis, 2nd ed.; Harper and Row: New York, NY, USA, 1967. [Google Scholar]
- U.S. Army Corps of Engineers, USACE. Hydrology Modelling System HEC-HMS, Version 4.2.1; U.S. Army Corps of Engineers: Davis, CA, USA, 2017; Available online: https://www.hec.usace.army.mil/software/hec-hms/documentation/HEC-HMS_ReleaseNotes421.pdf (accessed on 10 December 2018).
- U.S. Army Corps of Engineers. USACE. HEC-RAS River Analysis System, Version 5.0; U.S. Army Corps of Engineers: Davis, CA, USA, 2016; Available online: https://www.hec.usace.army.mil/software/hec-ras/documentation/HEC-RAS%205.0%20Reference%20Manual.pdf (accessed on 10 December 2018).
- Chuenchooklin, S.; Taweepong, S.; Pangnakorn, U. Rainfall-runoff Models and Flood Mapping in a Catchment of the Upper Nan Basin. Thailand. J. Water Resour. Hydraul. Eng.
**2015**, 4, 293–302. [Google Scholar] [CrossRef] - Chow, V.T.; Maidment, D.R.; Mays, L.W. Applied Hydrology. 1988. McGraw-Hill. Singapore. Available online: http://ponce.sdsu.edu/Applied_Hydrology_Chow_1988.pdf (accessed on 10 December 2018).
- Kundzewicz, Z.W.; Kanae, S.; Seneviratne, S.I.; Handmer, J.; Nicholls, N.; Peduzzi, P.; Mechler, R.; Bouwer, L.M.; Arnell, N.; Mach, K.; et al. Flood risk and climate change: Global and regional perspectives. Hydrol. Sci. J.
**2014**, 59, 1–28. [Google Scholar] [CrossRef] [Green Version]

**Figure 1.**Number of deaths due to the fact of floods worldwide from 1900 to 2020 [3].

**Figure 3.**Nan River near Nan Province and its municipality area for hydrodynamic modeling (HEC-RAS)

**.**

**Figure 4.**Methodology and computational procedure for calculation of the flood hazard, flood damage vulnerability, and flood risk at the household level in the municipal study area.

**Figure 5.**Configuration of the HEC-HMS and HEC-RAS models for the upstream sub-basins of the Nan River and the Nan municipal area.

**Figure 6.**Results of the HEC-HMS model’s calibration from 2006 to 2011 and verification from 2012 to 2017 at Tha Wang Pha station (N64) and Mueang Nan station (N1).

**Figure 7.**Results of the HEC-RAS model’s calibration of flood events in 2006 and 2011 and its verification in 2013 and 2016 at Mueang Nan station (N1), Nan River.

Hazard Parameters | Hazard Level | Hazard Index |
---|---|---|

Flood Duration, T (h) | FHIT | |

T ≤ 60 | Very Low | 1 |

60 < T ≤ 100 | Low | 2 |

100 < T ≤ 120 | Medium | 3 |

120 < T ≤ 140 | High | 4 |

140 < T | Very High | 5 |

Flood Depth, D (m) | FHID | |

D ≤ 0.60 | Very Low | 1 |

0.60 < D ≤ 2.00 | Low | 2 |

2.00 < D ≤ 2.25 | Medium | 3 |

2.25 < D ≤ 2.50 | High | 4 |

2.50 < D | Very High | 5 |

Flood Velocity, V (m/s) | FHIV | |

V ≤ 0.10 | Very Low | 1 |

0.10 < V ≤ 0.50 | Low | 2 |

0.50 < V ≤ 0.60 | Medium | 3 |

0.60 < V ≤ 0.70 | High | 4 |

0.70 < V | Very High | 5 |

**Table 2.**Classification of population vulnerability, household vulnerability, vulnerability levels, and indices.

Range of Population Vulnerability, VI _{pop} (Person/km^{2}) | Vulnerability Level | FVI_{pop} Index |
---|---|---|

VI_{pop} ≤ 1000 | Very Low | 1 |

1000 < VI_{pop} ≤ 2000 | Low | 2 |

2000 < VI_{pop} ≤ 3000 | Medium | 3 |

3000 <VI_{pop} ≤ 4000 | High | 4 |

4000 < VI_{pop} | Very High | 5 |

Range of HouseholdVulnerability, VI_{hh}(Equation (5)) | VulnerabilityLevel | FVI_{hh} Index |

VI_{hh} ≤ 22.6 | Very Low | 1 |

22.6 < VI_{hh} ≤ 29.2 | Low | 2 |

29.2 < VI_{hh} ≤ 35.8 | Medium | 3 |

35.8 < VI_{hh} ≤ 42.4 | High | 4 |

42.4 < VI_{hh} | Very High | 5 |

**Table 3.**Major contributing factors of household vulnerability and their components in Phumin-Thali village.

Major Con-tributing Factors F_{i} in Equations (5) and (6) | Weights W_{i} of F_{i} in Equations (5) and (6) | Components C _{i} of Factor F_{i} in Equation (7) & Definitions | Classification of Component C_{i} in Equation (7) into j Classes and Their Ranges | % of Total Collected Samples | Impact Score K_{j} of Class j in % in Equation (8) | Weights θ _{i} of C_{i} in Equations (8) and (9) | Remarks on Definition of Each Class of C_{i} |
---|---|---|---|---|---|---|---|

F1 = Sensitivity | W_{1} = 0.26 | C_{1} = family size | Class 1 = family members > 5; Class 2 = members between 3–5; Class 3 = less than 3 | Q1 = 16.7 Q2 = 33.3 Q3 = 66.7 | 100 67 33 | θ_{1} = 0.386 | Larger family is more sensitive to flooding |

C_{2} = gender of householder | Class 1 = female; Class 2 = male | Q1 = 33.3 Q2 = 66.7 | 100 40 | θ_{2} = 0.193 | Female is more sensitive to flooding than male | ||

C_{3} = health of householder | Class 1 = very poor; Class 2 = poor; Class 3 = good; Class 4 = very good | Q1 = 0 Q2 = 16.7 Q3 = 50.0 Q4 = 33.3 | 100 75 50 25 | θ_{3} = 0.129 | Very poor health person is most sensitive to flooding | ||

C_{4} = land use type | Class 1 = houses + orchards; Class 2 = houses + shops; Class 3 = houses only | Q1 = 0 Q2 = 50 Q3 = 50 | 100 67 33 | θ_{4} = 0.055 | House + orchard land is most vulnerable and most sensitive to flooding | ||

C_{5} = household damages, THB | Class 1 = damages >10,000; Class 2 = damages 5000–10,000; Class 3 = damages <5000; Class 4 = no damage | Q1 = 0 Q2 = 33.3 Q3 = 16.7 Q4 = 50.0 | 100 75 50 25 | θ_{5} = 0.096 | Household with higher damage potential is more sensitive to flooding | ||

C_{6} = public damages (THB/m^{2}) | Class 1 = damages >750; Class 2 = damages 750–501; Class 3 = 501–250; Class 4 = less than 250 | Q1 = 0 Q2 = 100 Q3 = 0 Q4 = 0 | 100 75 50 25 | θ_{6} = 0.077 | Public property of higher damages is more sensitive to flooding | ||

C_{7} = household ownership | Class 1 = owner; Class 2 = tenant | Q1 = 66.7 Q2 = 33.3 | 100 0 | θ_{7} = 0.064 | Owners are more sensitive to flooding than tenants | ||

F2 = Adaptive Capacity | W_{2} = 0.11 | C_{1} = education level of householder | Class 1 = university; Class 2 = higher secondary; Class 3 = secondary; Class 4 = primary; Class 5 = illiterate | Q1 = 16.7 Q2 = 33.3 Q3 = 0 Q4 = 50.0 Q5 = 0 | 100 80 60 40 20 | θ_{1} = 0.129 | Higher educated person has more adaptive capacity to flooding |

C_{2} = type of employment of householder | Class 1 = Govt. officer; Class 2 = private worker; Class 3 = agriculture; Class 4 = daily wage Class 5 = unemployed | Q1 = 16.7 Q2 = 0 Q3 = 0 Q4 = 83.3 Q5 = 0 | 100 80 60 40 20 | θ_{2} = 0.096 | Government officer has highest adaptive capacity to flooding | ||

C_{3} = householder income (THB/month) | Class 1 = income >25,000; Class 2 = income 25,000–20,001; Class 3 = 20,000–15,001; Class 4 = 15,000–10,001; Class 5 = 10,000–5001; Class 6 = less than 5000 | Q1 = 16.7 Q2 = 16.7 Q3 = 0 Q4 = 16.7 Q5 = 16.7 Q6 = 33.3 | 100 83 67 50 33 16.7 | θ_{3} = 0.386 | Householder with higher income has more adaptive capacity to flooding | ||

C_{4} = household saving deposit | Class 1 = yes; Class 2 = no | Q1 = 100 Q2 = 0 | 100 0 | θ_{4} = 0.193 | Household with saving deposit has more adaptive capacity | ||

C_{5} = flood insurance | Class 1 = yes; Class 2 = no | Q1 = 0 Q2 = 100 | 100 0 | θ_{5} = 0.055 | Household with flood insurance has more adaptive capacity | ||

C_{6} = land price(THB/m ^{2}) | Class 1 = land price >7500 Class 2 = 7500–5001; Class 3 = 5000–2501; Class 4 = 1–2500 | Q1 = 0 Q2 = 0 Q3 = 100 Q4 = 0 | 100 75 50 25 | θ_{6} = 0.064 | Household in high price land is richer and has more adaptive capacity | ||

C_{7} = flood notification | Class 1 = yes; Class 2 = no | Q1 = 100 Q2 = 0 | 100 0 | θ_{7} = 0.077 | People with flood notification have more adaptive capacity | ||

F3 = Exposure | W_{3} = 0.63 | C_{1} = distance from river, m | Class 1 = distance <500, Class 2 = 501–1000; Class 3 = 1001–1500; Class 4 = 1501–2000; Class 5 = more than 2000 | Q1 = 100 Q2 = 0 Q3 = 0 Q4 = 0 Q5 = 0 | 100 80 60 40 20 | θ_{1} = 0.408 | Area at shorter distance to river is more exposed to flooding |

C_{2} = ground elevation, m | Class 1 = elev. 190–195, Class 2 = 196–200, Class 3 = 201–205, Class 4 = 206–210, Class 5 = elev.211–215 | Q1 = 100 Q2 = 0 Q3 = 0 Q4 = 0 Q5 = 0 | 100 80 60 40 20 | θ_{2} = 0.204 | Area with lower elevation is more exposed to flooding | ||

C3 = inundation depth in 2011, m | Class 1 = depth >2.0; Class 2 = 1.1- 2.0; Class 3 = 0.1–1.0; Class 4 = less than 0.1 | Q1 = 0 Q2 = 83.3 Q3 = 16.7 Q4 = 0 | 100 75 50 25 | θ_{3} = 0.102 | Area having larger inundation depth is more exposed to flooding | ||

C4 = flood velocity in 2011, ms^{−1} | Class 1 = velocity >2.0; Class 2 = 1.1–2.0; Class 3 = 0.1–1.0; Class 4 = less than 0.1 | Q1 = 16.7 Q2 = 50 Q3 = 33.3 Q4 = 0 | 100 75 50 25 | θ_{4} = 0.082 | Area having higher flood flow velocity is more exposed to flooding | ||

C5 = duration of inundation in 2011, days | Class 1 = duration > 7; Class 2 = 3.5–7; Class 3 = 0.5–3.5; Class 4 = less than 0.5 | Q1 = 0 Q2 = 50 Q3 = 50 Q4 = 0 | 100 75 50 25 | θ_{5} = 0.136 | Area having longer flood duration is more exposed to flooding | ||

C6 = number of flooding events in 2011 | Class 1 = 3 flooding events or more; Class 2 = 2 flooding events; Class 3 = 1 flooding event; Class 4 = none | Q1 = 0 Q2 = 33.3 Q3 = 66.7 Q4 = 0 | 100 75 50 25 | θ_{6} = 0.068 | Area having more frequent flooding is more vulnerable to flooding |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Tingsanchali, T.; Promping, T.
Comprehensive Assessment of Flood Hazard, Vulnerability, and Flood Risk at the Household Level in a Municipality Area: A Case Study of Nan Province, Thailand. *Water* **2022**, *14*, 161.
https://doi.org/10.3390/w14020161

**AMA Style**

Tingsanchali T, Promping T.
Comprehensive Assessment of Flood Hazard, Vulnerability, and Flood Risk at the Household Level in a Municipality Area: A Case Study of Nan Province, Thailand. *Water*. 2022; 14(2):161.
https://doi.org/10.3390/w14020161

**Chicago/Turabian Style**

Tingsanchali, Tawatchai, and Thanasit Promping.
2022. "Comprehensive Assessment of Flood Hazard, Vulnerability, and Flood Risk at the Household Level in a Municipality Area: A Case Study of Nan Province, Thailand" *Water* 14, no. 2: 161.
https://doi.org/10.3390/w14020161