# The Effect of Multi-Source DEM Accuracy on the Optimal Catchment Area Threshold

^{1}

^{2}

^{3}

^{*}

## Abstract

**:**

^{2}, which has the highest coincidence with the actual river network. In summary, using the SRTM V4.1 DEM as the DEM data source is feasible to determine the optimal catchment area threshold in plain watersheds.

## 1. Introduction

## 2. Data Sources and Methods

#### 2.1. Overview of the Study Area

^{2}, with an average annual precipitation of 1024 mm and average annual evaporation of 781 mm, with rain and heat in the same season. The main geomorphological form is mainly plains and hills, with an elevation of 268~682 m and slight topographic undulation. The water system in the region is well developed, including the mainstream Minjiang River and its tributaries, Simeng River, Mao River, and Liquan River, forming a large irrigation area. The actual river network shape is determined, the corresponding digital river network is unique, and the fractal dimension of the water system is also uniquely determined. The accurate water system fractal dimension should be calculated from the actual river network to obtain. Based on the 1:50,000 topographic map obtained from the local hydrological management office and digitized by ARCGIS map, the actual river network in the water source area of Tongjiyan was extracted, as shown in Figure 1.

#### 2.2. Data Sources and Pre-Processing

#### 2.3. River Network Extraction Method

#### 2.4. Box Dimension Method

#### 2.5. River Network Hedge Differential Evaluation System

_{i}is the polygon area produced by the superposition of two river networks; M is the total area of the watershed. The judgment criteria [27] are as follows: when the set difference is less than 2%, the river network has a high degree of coincidence; when the set difference is between 2% and 3%, the degree of coincidence is more satisfactory; when the set difference is greater than 3%, the degree of coincidence is poor.

## 3. Results and Analysis

#### 3.1. Calculation Results and Analysis of Box Dimension Method

^{2}were calculated using Equation (1). The data are shown in Table 2.

^{2}for the 30 m accuracy and 0.0081 km

^{2}for the 90 m accuracy. In the gentle fluctuation stage, the change in fractal dimension changes smoothly as the catchment area threshold increases in the range of 0.450–3.600 km

^{2}. Decreasing by 0.02 for 90 m SRTMV4.1, 0.0296 for 90 m HYDRO1K, 0.0382 for 90 m ASTER GDEM, and 0.0213 for 30 m ASTER GDEM. The R

^{2}of their fits all reach above 0.9990. Tending to 1 stage, the fractal dimension is basically constant with the increase in the catchment area threshold, and the value tends to 1.

#### 3.2. Selection of Optimal Catchment Area Thresholds for Watersheds

^{2}. Therefore, 4.05 km

^{2}is the optimal catchment threshold for the water source area of Tongjiyan. However, the fractal dimensions calculated by the three DEM data are very close to each other, so the reasonableness of the optimal catchment area threshold is evaluated.

#### 3.3. Evaluation of the Reasonableness of the Optimal Catchment Area Threshold

#### 3.3.1. Similarity of River Network

^{2}, and the catchment area is 772.08 km

^{2}. The river network overlap calculated by the ASTER GDEM is 2.306%, which is satisfactory and good; the river network overlap calculated by the HYDRO1K is 2.763%, which is poor. The river network overlap difference calculated by SRTM V4.1 is 0.39 times of ASTER GDEM and 0.32 times of HYDRO1K. The extracted digital river network is the most accurate. It best matches the actual river network, which effectively verifies the reasonableness of the optimal catchment area threshold in the water source area of Tongjiyan.

#### 3.3.2. Response of Terrain Features to DEM Accuracy

## 4. Discussion

## 5. Conclusions

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Acknowledgments

## Conflicts of Interest

## References

- Sun, A.L.; Yu, Z.B.; Yang, C.G.; Gu, H. Impact factors of contribution area threshold in extracting drainage network for rivers in China. J. Hydraul. Eng.
**2013**, 44, 901–908. [Google Scholar] - Dou, M.; Yu, L.; Jin, M. Study on relationship between box dimension and connectivity of river system in Huaihe River basin. J. Hydraul. Eng.
**2019**, 50, 670–678. [Google Scholar] - Wang, G.Z.; Li, Z.Y.; Zhang, J.Y.; Qu, J.G.; Hao, J.; Wang, Z.M. Influence of Grid Sizes’ change on the fractal dimension of water systems. J. Irrig. Drain.
**2017**, 36, 215–217. [Google Scholar] - Yang, J.L. Digital determination of catchment area threshold based on fractal dimension. Sci. Surv. Mapp.
**2011**, 36, 33–34. [Google Scholar] - O’Callaghan, J.F.; David, M. The extraction of drainage networks from digital elevation data. Comput. Vis. Graph. Image Process.
**1984**, 27, 323–344. [Google Scholar] - Lin, F.; Chen, X.W.; Wang, L. Calculation of the fractal dimension in Jiulongjiang River basin based on DEM. J. Water Resour. Water Eng.
**2009**, 20, 29–32. [Google Scholar] - Wang, L.; Chen, X.W. Study on relationship between extracted river network and fractal dimension based on DEM. J. Geo-Inf. Sci.
**2007**, 4, 133–137. [Google Scholar] - Li, W.T.; Wang, C.; Li, P.; Gu, L.W. Influences of DEM generalization methods on extracting accuracy of hydrological structure information. Bull. Soil Water Conserv.
**2012**, 32, 168–171. [Google Scholar] - Zheng, N.J. Fractal analysis of drainage networks based on box dimension. China Rural. Water Hydropower
**2020**, 5, 33–38. [Google Scholar] - Zhu, J.W. On the methods of DEM construction based on stereo image matching data. Bull. Surv. Mapp.
**2008**, 11, 38–40. [Google Scholar] - He, D.L.; Wen, J.G.; Xiao, Q.; You, D.Q.; Wu, X.D.; Lin, X.W.; Wu, S.B. An accuracy assessment method of DEM upscaling based on energy factor. Geomat. Inf. Sci. Wuhan Univ.
**2019**, 44, 570–577. [Google Scholar] - He, L.; Sun, H.; Huang, Y.L.; Huang, X.Y. Study on resampling methods of MODIS 1B data. Remote Sens. Informa.
**2007**, 3, 39–43. [Google Scholar] - Jiang, G.X.; Xie, Y.L.; Gao, Z.Y.; Zhou, P. Evaluation on elevation accuracy of commonly used DEM in five typical areas of China. Res. Soil Water Conserv.
**2020**, 27, 72–80. [Google Scholar] - Chen, G.H.; Li, Z.W.; Xu, X.W.; Sun, H.; Ha, G.; Guo, P.; Su, P.; Yuan, Z.; Li, T. Co-seismic surface deformation and late Quaternary accumulated displacement along the seismogenic fault of the 2021 Madoi M7.4 earthquake and their implications for regional tectonics. Chin. J. Geophys.
**2022**, 65, 2984–3005. [Google Scholar] - Tang, G.A.; Na, J.M.; Cheng, W.M. Progress of DEM terrain analysis on regional geomorphology in China. Acta Geod. Cartogr. Sin.
**2017**, 46, 1570–1591. [Google Scholar] - Saied, P. Coupling of DEM and remote-sensing-based approaches for semi-automated detection of regional geostructural features in Zagros Mountain, Iran. Arab. J. Geosci.
**2013**, 6, 91–99. [Google Scholar] - Saied, P.; Jonathan, L.; Michael, C. Use of LiDAR-derived DEM and a stream length gradient index approach to investigation of landslides in Zagros Mountains, Iran. Geocarto Int.
**2018**, 33, 912–926. [Google Scholar] - Parizi, E. Application of Unmanned Aerial Vehicle DEM in flood modeling and comparison with global DEMs: Case study of Atrak River Basin, Iran. J. Environ. Manag.
**2022**, 317, 115492. [Google Scholar] [CrossRef] - Ahmad, S.; Elnaz, P. A strategy to assess the uncertainty of a climate change impact on extreme hydrological events in the semi-arid Dehbar catchment in Iran. Theor. Appl. Climatol.
**2020**, 139, 389–402. [Google Scholar] - Srishti, G.; Arnab, B.; Rajendra, S. Modelling potential impact of climate change and uncertainty on streamflow projections: A case study. J. Water Clim. Chang.
**2020**, 12, 384–400. [Google Scholar] - Ehteram, M. Reservoir operation based on evolutionary algorithms and multi-criteria decision making under climate change and uncertainty. J. Hydroinformatics
**2018**, 20, 332–355. [Google Scholar] [CrossRef] - Cai, L.Y.; Tang, G.A.; Xiong, L.Y.; Kamila, J.L.; Xie, T. An analysis on fractal characteristics of typical landform patterns in Northern Shanxi Loess Plateau based on DEM. Bull. Soil Water Conserv.
**2014**, 34, 141–144. [Google Scholar] - Lu, Z.X.; Zou, S.B.; Xiao, H.L.; Yin, Z.L.; Ruan, H.W.; Gao, H.S. How to determine drainage area threshold in alpine regions of upper reaches of Heihe River. J. Glaciol. Geocryol.
**2015**, 37, 493–499. [Google Scholar] - Zhang, P.J.; Meng, X.M.; Zhou, H.; Liu, D.F. Applicability analysis of different flow accumulation threshold determination methods in loess eroded area. Water Resour. Power
**2019**, 37, 26–29. [Google Scholar] - Sadegh, A.M.; Reza, N.M.; Abrishamchi, A. Increasing risk of meteorological drought in the Lake Urmia basin under climate change: Introducing the precipitation–temperature deciles index. J. Hydrol.
**2020**, 592, 125586. [Google Scholar] - Ömer, E.; Mehmet, C.D.; Martijn, J.B. Effect of data length, spin-up period and spatial model resolution on fully distributed hydrological model calibration in the Moselle basin. Hydrol. Sci. J.
**2022**, 67, 759–772. [Google Scholar] - Song, H.H. Research on SRTM Void-filling and precision detection of SRTM data in China. Master’s Thesis, China University of Geosciences, Wuhan, China, 2018. [Google Scholar]
- Zheng, W.; Gao, H.L.; Zhang, M.H.; Zhang, Q.F. Study on methods of determining optimal catchment area threshold in Zhifanggou catchment. Res. Soil Water Conserv.
**2020**, 44, 901–908. [Google Scholar] - Wijaya, O.T.; Yang, T.-H. A novel hybrid approach based on cellular automata and a digital elevation model for rapid flood assessment. Water
**2012**, 13, 1311. [Google Scholar] [CrossRef] - Liu, Y.; Zhou, M.C.; Chen, Z.J.; Li, S.W. Comparison of drainage network extraction from different DEM data sources: A case study of Hanjiang River basin. Sci. Geogr. Sin.
**2012**, 32, 1112–1118. [Google Scholar] - Wang, F.; Wang, C.M. Research for the influences of DEM resolution on topographical factors—A case study of Mengyin County. Res. Soil Water Conserv.
**2009**, 16, 225–229. [Google Scholar] - Ma, Y.M.; Zhang, L.H.; Zhang, K.; Zhu, Z.R.; Wu, Z.F. Accuracy analysis of watershed system extraction based on the SWAT model and multi-source DEM data. Geogr. Geo.-Inf. Sci.
**2019**, 21, 1527–1537. [Google Scholar]

**Figure 6.**Effect of overlaying 3 kinds of DEM data sources. (

**a**) SRTM V4.1 DEM. (

**b**) HYDRO1K DEM. (

**c**) ASTER GDEM.

**Figure 7.**Slope maps extracted from three types of DEM data sources. (

**a**) SRTM V4.1 DEM. (

**b**) HYDRO1KDEM. (

**c**) ASTER GDEM.

Data Sources | Publishing Organization | Data Range | Access Channels | Horizontal Resolution | Vertical Accuracy |
---|---|---|---|---|---|

ASTER GDEM | METI/NASA | 83° N~83° S | http://www.gscloud.cn (accessed on 14 March 2022) | 30 m | 20 m |

SRTM V4.1 DEM | USGS/NASA | 60° N~56° S | http://www.gscloud.cn (accessed on 14 March 2022) | 90 m | Nominal: 16 m; Actual: 9 m |

HYDRO1K DEM | USGS/EROS | Global | https://lta.cr.usgs.gov/ HYDRO1K (accessed on 14 March 2022) | 1000 m | 30 m |

Raster | 90 m SRTM GDEM | 90 m HYDRO1K DEM | 90 m ASTER GDEM | 30 m ASTER GDEM | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|

Threshold /km ^{2} | D | R^{2} | Threshold /km ^{2} | D | R^{2} | Threshold /km ^{2} | D | R^{2} | Threshold /km ^{2} | D | R^{2} | |

1 | 0.0081 | 1.3948 | 0.9861 | 0.0081 | 1.7845 | 0.9990 | 0.0081 | 1.5866 | 0.9991 | 0.0090 | 1.9091 | 0.9992 |

500 | 0.4050 | 1.0450 | 0.9996 | 0.4050 | 1.0548 | 0.9991 | 0.4050 | 1.0080 | 0.9993 | 0.4500 | 1.0428 | 0.9991 |

1000 | 0.8100 | 1.0294 | 0.9999 | 0.8100 | 1.0310 | 0.9990 | 0.8100 | 1.0296 | 0.9999 | 0.9000 | 1.0215 | 0.9995 |

2000 | 1.6200 | 1.0268 | 0.9999 | 1.6200 | 1.0301 | 0.9993 | 1.6200 | 1.0268 | 0.9999 | 1.8000 | 1.0343 | 0.9999 |

3000 | 2.4300 | 1.0271 | 0.9999 | 2.4300 | 1.0220 | 0.9998 | 2.4300 | 1.0001 | 0.9999 | 2.7000 | 1.0281 | 0.9999 |

4000 | 3.2400 | 1.0250 | 0.9999 | 3.2400 | 1.0252 | 0.9999 | 3.2400 | 0.9914 | 0.9999 | 3.6000 | 1.0252 | 0.9999 |

5000 | 4.0500 | 1.0245 | 0.9999 | 4.0500 | 1.0278 | 0.9999 | 4.0500 | 1.0249 | 0.9999 | 4.5000 | 1.0239 | 0.9999 |

6000 | 4.8600 | 1.0242 | 0.9999 | 4.8600 | 1.0251 | 0.9999 | 4.8600 | 1.0242 | 0.9999 | 5.4000 | 1.0240 | 0.9999 |

7000 | 5.6700 | 1.0239 | 0.9999 | 5.6700 | 1.0241 | 0.9999 | 5.6700 | 1.0237 | 0.9999 | 6.3000 | 1.0228 | 0.9999 |

8000 | 6.4800 | 1.0240 | 0.9999 | 6.4800 | 1.0237 | 0.9999 | 6.4800 | 1.0235 | 0.9999 | 7.2000 | 1.0230 | 0.9999 |

Data Sources of DEM | Fractal Dimension D | Optimal Catchment Area Threshold/km^{2} |
---|---|---|

Actual River Network | 1.0245 | / |

SRTM V4.1 DEM | 1.0245 | 4.05 |

HYDRO1K DEM | 1.0241 | 5.67 |

ASTER GDEM | 1.0242 | 4.86 |

Data Sources of DEM | Polygon Area /km ^{2} | Watershed Area /km ^{2} | River Network Overlap Difference/% |
---|---|---|---|

SRTM V4.1 DEM | 6.93 | 772.08 | 0.898 |

HYDRO1K DEM | 21.30 | 770.89 | 2.763 |

ASTER GDEM | 17.79 | 771.39 | 2.306 |

Data of DEM | Maximum Slope/° | Average Slope/° | Standard Deviation of Slope | Maximum Elevation/m | Minimum Elevation/m | Undulation/m |
---|---|---|---|---|---|---|

90 m SRTM DEM | 71.10 | 5.40 | 5.04 | 680 | 268 | 412 |

90 m HYDRO1K DEM | 70.89 | 5.01 | 4.64 | 669 | 285 | 384 |

90 m ASTER GDEM | 69.83 | 5.09 | 4.79 | 675 | 289 | 386 |

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |

© 2023 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**

Wu, H.; Liu, X.; Li, Q.; Hu, X.; Li, H.
The Effect of Multi-Source DEM Accuracy on the Optimal Catchment Area Threshold. *Water* **2023**, *15*, 209.
https://doi.org/10.3390/w15010209

**AMA Style**

Wu H, Liu X, Li Q, Hu X, Li H.
The Effect of Multi-Source DEM Accuracy on the Optimal Catchment Area Threshold. *Water*. 2023; 15(1):209.
https://doi.org/10.3390/w15010209

**Chicago/Turabian Style**

Wu, Honggang, Xueying Liu, Qiang Li, Xiujun Hu, and Hongbo Li.
2023. "The Effect of Multi-Source DEM Accuracy on the Optimal Catchment Area Threshold" *Water* 15, no. 1: 209.
https://doi.org/10.3390/w15010209