Neighbor Interpolation Research Articles

How does high resolution topography affect flood simulation at watershed scale: A case study in a small watershed based on tilt photography

The data source and resolution of DEM will affect the extraction of watershed features and flood simulation precision. In this paper, based on the 7 cm resolution DEM of the Liulin watershed obtained by tilt photography technology, the nearest neighbor interpolation, bilinear interpolation, and cubic convolution methods were used to resample DEMs with 1m and 30m resolution, and 6 resampled DEMs were obtained. The Aster 30m resolution DEM was downloaded from the geospatial data cloud. The watershed features were extracted from the 7 DEMs and compared. Based on these DEMs, the HEC-HMS model was established. 20 flood events were selected in the Liulin watershed for model calibration and verification. The results showed that the resampling method has little effect on the topographic features, but there were some differences between the resampling and downloaded DEM topographic features. The flood simulation results of the 1m resolution DEM-based HEC-HMS model are slightly better than that of the 30m resolution DEM-based model. The resampled 30m resolution DEM-based model performed identically with Aster 30m resolution DEM-based model. Therefore, an Aster 30m DEM can be used for flood simulation in a watershed that lacks measured high-precision topographic data.

Quantum image scaling with applications to image steganography and fusion

A quantum hue, saturation, and lightness model is proposed in which a triple qubit sequence (QHTS) is encoded and used as a data model for the implementation of quantum image scaling. The preparation and retrieval of QHTS images is presented, in which only q+2 qubits (where q is the bit depth) are required to encode color information while retaining relevant HSL image features and operability. A conventional nearest neighbor interpolation was adopted to implement quantum image up-scaling and down-scaling operations, from which two other scaling applications were developed. One such technique is a form of quantum steganography based on end-to-end encryption, which provides high capacity while ensuring the security of carrier images and secret messages. The other is a spatial remote sensing image fusion algorithm, based on QHTS images, which pioneers quantum pseudocolor composites of multi-spectral and panchromatic images. Simulation experiments demonstrated the proposed methodology provides an embedding capacity more than double that of existing quantum image steganography algorithms. In addition, a complexity analysis demonstrated the efficiency of the two proposed quantum image scaling applications, which take full advantage of quantum parallelism.

Research on High-Resolution Reconstruction of Marine Environmental Parameters Using Deep Learning Model

The analysis of marine environmental parameters plays a significant role in various aspects, including sea surface target detection, the monitoring of the marine ecological environment, marine meteorology and disaster forecasting, and the monitoring of internal waves in the ocean. In particular, for sea surface target detection, the accurate and high-resolution input of marine environmental parameters is crucial for multi-scale sea surface modeling and the prediction of sea clutter characteristics. In this paper, based on the low-resolution wind speed, significant wave height, and wave period data provided by ECMWF for the surrounding seas of China (specified latitude and longitude range), a deep learning model based on a residual structure is proposed. By introducing an attention module, the model effectively addresses the poor modeling performance of traditional methods like nearest neighbor interpolation and linear interpolation at the edge positions in the image. Experimental results demonstrate that with the proposed approach, when the spatial resolution of wind speed increases from 0.5° to 0.25°, the results achieve a mean square error (MSE) of 0.713, a peak signal-to-noise ratio (PSNR) of 49.598, and a structural similarity index measure (SSIM) of 0.981. When the spatial resolution of the significant wave height increases from 1° to 0.5°, the results achieve a MSE of 1.319, a PSNR of 46.928, and an SSIM of 0.957. When the spatial resolution of the wave period increases from 1° to 0.5°, the results achieve a MSE of 2.299, a PSNR of 44.515, and an SSIM of 0.940. The proposed method can generate high-resolution marine environmental parameter data for the surrounding seas of China at any given moment, providing data support for subsequent sea surface modeling and for the prediction of sea clutter characteristics.

Evaluating imaging uncertainty associated with the near surface and added value of vertical arrays using Bayesian seismic refraction tomography

AbstractTraditional tomographic methods do not consider the uncertainties associated with near-surface velocities and static corrections and provide a deterministic solution to the estimation problem. However, these uncertainties significantly affect structural mapping and interpretation of seismic imaging results. On the other hand, Bayesian first-arrival tomography provides multiple near-surface models that fit observed traveltimes equally well and enable the study of potential solution distributions. We demonstrate this approach on a complex synthetic near-surface model, representative of arid environments, to quantify associated velocity and statics uncertainties. We evaluate two different parameterizations for subsurface velocities in the context of near-surface Bayesian tomography: Voronoi tessellation with natural neighbor interpolation and the more conventional Delaunay triangulation with linear interpolation. Our analysis shows that the Voronoi cell parameterization with natural neighbor interpolation is more appropriate for this problem. Finally, the new approach is applied to compare two alternative acquisition geometries comprising conventional surface receivers and surface receivers augmented with vertical receiver arrays. The results demonstrate that adding vertical receiver arrays to conventional surface receivers can significantly reduce the near-surface velocity uncertainty and thus increases the accuracy of the seismic imaging results. Furthermore, the study shows that Bayesian tomography can be used as a tool for evaluating different source and receiver geometries during the acquisition design stage.

Multiband 2D InSAR deformation models with error estimates from natural neighbour interpolation: Case study in the Latrobe Valley, Australia

A long term objective for Australia is to develop a continent-wide 2D deformation model based on satellite interferometric synthetic aperture radar (InSAR). Nation-wide InSAR processing has been demonstrated in European countries and Japan. However, there are some key differences to the Australian situation for a 2D model (i.e., vertical and east–west surface movement rates) where SAR data from ascending and descending satellite orbits are needed. The primary difference is that while ESA’s Sentinel-1 ascending and descending satellite observations underpin 2D models in other parts of the world, the Australian continent is only covered by descending Sentinel-1 SAR data in most areas. Combined with Australia’s large size of 7.63 million km2, this means a different approach is needed to realise 2D deformation models. In this article, we propose the novel approach of (1) combining supplementary multiband SAR (ascending X-band and L-band) with the descending C-band Sentinel-1 data, and (2) using natural neighbour (NN) interpolation as a computationally efficient interpolation method for combining the multiband InSAR data on common grids prior to 2D model computation. This is an alternative to the computationally intensive Ordinary Kriging (OK) approach. A disadvantage of NN compared to OK is that it does not compute an error estimate, which we address by deriving an error estimate for the NN interpolation so that errors are provided for each 2D rate to aid with their correct interpretation. To demonstrate our new approach, we compute experimental 2D deformation models that propagate the full error budget from InSAR, a GNSS connection to ITRF2014 and NN interpolation into an error grid in the Latrobe Valley in southeast Australia. The significance of our approach is that it provides a suitable combination of methods that may be used to develop 2D continental deformation models for Australia’s unique circumstances. Our results demonstrate the accuracy of NN and OK interpolation are at the same level (within error), and the error estimates from both interpolation techniques are also comparable: X band (TerraSAR-X) ∼ 1 mm/yr; C band (Sentinel-1) ∼ 2.5 mm/yr, and L band (ALOS PALSAR1) 4 mm/yr. For the first time, we reveal 2D deformation in the Latrobe Valley region showing horizontal east–west deformation in the area surrounding an above-ground mine and reaching 20 mm/yr, with vertical subsidence rates reaching −40 mm/yr.

Exploring modern bathymetry: A comprehensive review of data acquisition devices, model accuracy, and interpolation techniques for enhanced underwater mapping

Building a high-precision bathymetry digital elevation model is essential for navigation planning, marine and lake resource planning, port construction, and underwater archaeological projects. However, existing bathymetry methods have yet to be effectively and comparatively analyzed. This paper comprehensively reviews state-of-the-art bathymetry methods, including data acquisition techniques, model accuracy, and interpolation algorithms for underwater terrain mapping. First, We assess the merits and drawbacks of novel data acquisition devices, such as single-beam/multi-beam echo sounders and light detection and ranging systems. After that, we analyze the accuracy of the ETOPO1, GEBCO_2022 and SRTM15 to provide valuable insights into their performance. Furthermore, we evaluate ANUDEM, Inverse Distance Weighting, Kriging and Nearest Neighbor interpolation algorithms in different underwater terrains by comparing their applicability, reliability, and accuracy in various underwater environments. Finally, we discuss the development trends and challenges in underwater bathymetry technology and offer a forward-looking perspective on the future of this essential field.

OpenPose based Smoking Gesture Recognition System using Artificial Neural Network

Smoking is an extremely important health problem in modern society. This study focuses on a method for preventing smoking in non-smoking areas, such as public places, as well as the development of an artificial neural network based smoking motion recognition system for more accurately recognizing smokers in such areas. In particular, we attempted to increase the rate of recognition of smoking behaviors using an OpenPose based algorithm and the accuracy of such recognition by additionally applying a hardware device for recognizing cigarette smoke. In addition, a preprocessing method for inputting a dataset into the proposed system is proposed. To improve the recognition performance, four types of dataset models were created, and the most suitable dataset model was selected experimentally. Based on this dataset model, test data were created and input into the proposed neural network based smoking behavior recognition system. In addition, the nearest neighbor interpolation method was selected experimentally as an image interpolation approach and applied to the image preprocessing. When applying experimental data based on learned data, the developed system showed a recognition rate of 70-75%, and the smoking recognition accuracy was increased through the addition of the hardware device.

Glass transition of amorphous polymeric materials informed by machine learning

The glass transition temperature (Tg) is used to determine thermophysical properties of polymer materials and is often considered one of the most important descriptors. Methods for predicting various physical properties of materials based on machine learning algorithms and key molecular descriptors are efficient and accurate. However, it still needs improvements because an overly complex model is less practical and difficult to generalize. In addition, obtaining a large number of samples to achieve accurate predictions remains a challenge due to the complex and lengthy experimental process. In this work, based on Tg of 100 polymers, we use a feature selection algorithm combining FeatureWiz and the least absolute shrinkage and selection operator to quickly select molecular descriptors that are minimally redundant and maximally relevant to Tg. The processed dataset is interpolated from the original dataset using the nearest neighbor interpolation algorithm to solve the data deficiency problem. Finally, the synthetic minority oversampling technique algorithm is used to solve the data imbalance problem. The augmented dataset is used to construct the extreme gradient boosting prediction model to achieve good prediction accuracy. The experimental results demonstrate the robustness of the proposed model and the accuracy of its prediction results.

Enhanced chip analysis with computed tomography for estimation of chip segmentation frequency

This work presents an enhancement of a method for analyzing metallic chips produced during turning processes in order to estimate the chip segmentation frequency. Therefore, the resulting 3D image of a computer tomography scan of the chip can be transformed virtually into a straight chip by describing the course of the chip in three-dimensional space by a trajectory. Along this trajectory, the chip point cloud is sliced in thin layers which can be transformed to a straight chip. To estimate the segment rate, the height course describing the segmentation of the chip is extracted on the serrated side of the point cloud by means of nearest neighbor interpolation. The resulting signals are compared with the measurements of the chip surface using an optical microscope, whereby only two straight chips could be analyzed. The comparison shows a nearly perfect match between the two signals and is confirmed by a Pearson correlation coefficient of over {86}% for both signals. Furthermore, a statistical comparison with the acoustic emissions of the turning process reveals satisfying similarities for the chip segmentation frequency with respect to the compression factor.

Analysis of Terrain Modelling Methods in the Coastal Zone

Geospatial data are increasingly used to model the terrain in the coastal zone, in particular in shallow waterbodies (with a depth of up to 1 m). In order to generate a terrain relief, it is important to choose a method for its modelling that will allow it to be accurately projected. Therefore, the aim of this publication is to analyze the terrain modelling methods in the coastal zone. For the purposes of the research, five most popular methods for terrain modelling were described: Inverse Distance Weighted (IDW), Modified Shepard's Method (MSM), Natural Neighbor Interpolation (NNI), kriging and spline. Each of the methods has been described in a uniform way in terms of: the essence of its operation, mathematical expression and application examples. The advantages and disadvantages of each of the methods for terrain modelling are also discussed. It should be stated that the choice of the method for the terrain modelling of a shallow waterbody is not unambiguous, as it depends on the type of data recorded during bathymetric and photogrammetric measurements.

The Impact of Missing Data on Heart Rate Variability Features: A Comparative Study of Interpolation Methods for Ambulatory Health Monitoring

ObjectivesHeart rate variability (HRV) is a valuable indicator of both physiological and psychological states. However, the accuracy of HRV measurements taken by wearable devices can be compromised by errors during transmission and acquisition. These errors can significantly affect HRV features and are not acceptable for precise HRV analysis used for medical diagnosis. This study aims to address this issue by investigating the effectiveness of four different interpolation methods (Nearest Neighbour - NN, Linear, Shape-preserving piecewise cubic Hermite - Pchip, and cubic spline) in tackling missing RR values in real-time HRV analysis. Materials and MethodsIn this study, HRV signals were obtained from Electrocardiograms (ECG) through automatic detection and manually corrected by a specialist, resulting in high-quality signals with no missing or ectopic peaks. To simulate low-quality data acquisition, values were iteratively deleted from each HRV analysis window. The deleted values were then replaced using four different interpolation methods. Time and frequency domain features were computed from both the original and reconstructed signals, and the Mean Absolute Percentage Error (MAPE) was used to compare these features. ResultsResults showed that as the percentage of missing values increased, some interpolation methods were more suitable for RR time-series with a greater number of missing data. Furthermore, the study suggests that the impact of interpolation on HRV features varied across different features and that SDNN is the least affected by interpolation. In the time domain, nearest neighbour interpolation gives the best results for up to 50% missing data. Beyond this threshold, it seems better not to use any interpolation for RMSSD. In the frequency domain however, the lowest errors of HRV feature estimation are obtained using linear or Pchip interpolation. To achieve maximum performance, it is recommended to adapt the interpolation method to both the percentage of missing values and the targeted HRV feature. ConclusionResults highlight the importance of choosing the appropriate interpolation method to accurately estimate HRV features in real-time analysis. Overall, the Pchip interpolation seems to yield the best results on most HRV features as it preserves the linear trend of the data while adding very light waves. The findings can be beneficial in the development of more precise and reliable wearable devices for real-time HRV monitoring.

One‐kilometre monthly air temperature and precipitation product over the Mongolian Plateau for 1950–2020

AbstractThe fragile ecological environment of the Mongolian Plateau (MP) is sensitive to climate change. It is necessary to fully understand the temperature and precipitation changes on the MP to ensure regional sustainable development. Most existing studies are conducted based on station observations which suffer from sparse distribution and limited spatial representativeness and cannot perfectly represent the climatic conditions of the whole region. By contrast, the long‐term, spatially continuous reanalysis products offer new opportunities for MP climate change research. However, the coarse spatial resolution and large uncertainties of reanalysis data limit their applicability to provide reliable climate information at finer scales. Within the delta downscaling framework, we downscale and correct the state‐of‐the‐art European Centre for Medium‐range Weather Forecasts ReAnalysis 5 land portion (ERA5‐Land) using a high‐resolution WorldClim reference climatology and generate a long‐term (1950–2020) 1‐km monthly air temperature and precipitation downscaled dataset. During the downscaling process, the results obtained by nearest neighbour, bilinear and bicubic interpolation methods are evaluated and compared. Air temperature result of bilinear method and precipitation result of nearest neighbour method, which have the best accuracy, are taken as the final downscaled data. The evaluation shows that the downscaled data has acceptable accuracy (overall MBE = −0.41°C, Corr = 0.997, RMSE = 1.51°C for air temperature, and MBE = 1.58 mm·month−1, Corr = 0.747, RMSE = 24.24 mm·month−1 for precipitation), and outperforms the original ERA5‐Land and widely used Climatic Research Unit (CRU) data. In addition, the downscaled data can also provide accurate and detailed temperature and precipitation change trends over the MP. With fine spatial resolution, long‐time span and good spatial continuity, the downscaled air temperature and precipitation will be a useful data source to explore climate change over the MP.

A tolerant sequential correction predictive energy management strategy of hybrid electric vehicles with adaptive mesh discretization

This paper establishes an adaptive correction predictive energy management strategy (EMS) to obtain optimal power distribution in the case of inaccurate prediction with excellent computational efficiency for parallel hybrid electric vehicle (HEV). Firstly, deep neural network (DNN) models are trained to forecast future speed sequence with different prediction horizons to prepare for online model predictive control (MPC) implementation. According to the degree of forecasting errors, a novel tolerant sequential correction algorithm as the solver of MPC strategy is selected in tuning mechanism instead of traditional dynamic programming (DP) to cope with the rough accumulated prediction. Besides, to compromise between fuel optimization and lower computation burden, statistical analysis of battery SOC variation distribution is gained from historical driving cycles' calculation through offline DP. Then, nearest neighbor interpolation model is fitted to generate optimal ranges of state variable in each step, which adaptive mesh discretization method is intelligently reducing state and control calculation grid's number. Numerical simulations demonstrate that the proposed tolerant sequential correction algorithm applied in MPC scheme with adaptive mesh discretization have yielded the favorable capability of the fuel economy in the consideration of inaccurately velocity prediction and excellent computational efficiency, which exhibits the practical adaptability in real driving routes.

A critical comparison of interpolation techniques for digital terrain modelling in mining

Digital modelling of a surface is crucial for Earth science and mining applications for many reasons. These days, high-tech digital representations are used to produce a high-fidelity topographic surface in the form of a digital terrain model (DTM). DTMs are created from 2D data-points collected by a variety of techniques such as traditional ground surveying, image processing, LiDAR, radar, and global positioning systems. At the points for which data is not available, the heights need to be interpolated or extrapolated from the points with measured elevations. There are several interpolation/extrapolation techniques available, which may be categorized based on criteria such as area size, accuracy or exactness of the surface, smoothness, continuity, and preciseness. In this paper we examine these DTM production methods and highlight their distinctive characteristics. Real data from a mine site is used, as a case study, to create DTMs using various interpolation techniques in Surfer® software. The significant variation in the resulting DTMs demonstrates that developing a DTM is not straightforward and it is important to choose the method carefully because the outcomes depend on the interpolation techniques used. In mining instances, where volume estimations are based on the produced DTM, this can have a significant impact. For our data-set, the natural neighbour interpolation method made the best predictions (R2 = 0.969, β = 0.98, P < 0.0001).

A large-scale MAGDM model based on SKNN and weighted clustering under incomplete information

Due to the complexity of multi-attribute group decision-making(MAGDM) problems, decision-makers frequently provide incomplete and hesitant evaluation information. How to make a valid decision based on incomplete and hesitant information is an important issue in MAGDM problems. Firstly, we novelly introduce a sequential K-nearest neighbor(SKNN) interpolation method to estimate the missing values based on the order of missing information proportions. When seeking the k-nearest neighbors of a decision-maker, an improved similarity measurement is proposed by considering the fuzzy-value similarity and the hesitation similarity between decision-makers simultaneously. Secondly, a K-means clustering algorithm based on attribute weighting is proposed to make the collective evaluation information more representative, and the attribute weights are obtained by constructing a mathematical model based on the minimum discrimination principle which can synthesize the subjective and objective weights information. Finally, an illustrative example and comparison analysis are demonstrated to show the validity and superiority of the proposed model.

Using observational mean‐flow data to drive large‐eddy simulations of a diurnal cycle at the SWiFT site

AbstractReproducing realistic date‐ and site‐specific unsteady wind conditions in large‐eddy simulations is becoming increasingly useful in wind energy. How to run a large‐eddy simulation to match observed conditions, however, remains an open research question. One approach that has received considerable attention is mesoscale‐to‐microscale coupling, in which information about the mesoscale weather, most commonly acquired from a mesoscale numerical weather model, is passed on to a microscale model. In this paper, we demonstrate how the recently developed profile‐assimilation technique, a form of mesoscale‐to‐microscale coupling, can be used to drive large‐eddy simulations solely based on observed mean‐flow profiles at a single location, bypassing the need for auxiliary mesoscale simulations. The new approach is evaluated for a diurnal cycle at the Scaled Wind Farm Technology site. Observed mean‐flow profiles from the ground up to a height of 2 km are reconstructed by aggregating measurements from multiple instruments, and gaps in the data are infilled with natural neighbor interpolation. We perform nine simulations using various forcing approaches to deal with data limitations. The results show that it is indeed possible to drive microscale large‐eddy simulation with observations using the profile‐assimilation technique, notwithstanding large gaps in virtual potential temperature measurements. However, profile assimilation with vertical smoothing of the error between the desired and actual profiles is required. Without that smoothing, the microscale simulations develop unrealistically high turbulence levels under many situations. Finally, we show that simulated mesoscale data can account for missing observations, although care is needed as both data sources are not necessarily compatible.

Weather Parameter Analysis Using Interpolation Methods

One of the key components in minimizing losses has been using rainfall forecasts at various temporal and spatial scales by businesses, governments, and farmers. Forecasts for other atmospheric characteristics are also important for agriculture. To maintain accurate weather forecasts, atmospheric procedures must be understood, measurements must be improved, and research and development must be continued. India’s Ministry of Earth Science has many programs and agencies, such as the India Meteorological Department (IMD), that strive to enhance forecasting. The efforts have culminated in fairly reliable predictions of rainfall patterns. This research aims to enhance prediction accuracy in different geographical areas ranging from weather subdivisions to agro-climate areas. We have used data that were obtained from the IMD, Pune. In the preprocessing technique session, we transformed data from the txt format to the csv format. We have used Matlab software for data analysis. Interpolation techniques have introduced detailed experimental position forecasts for the Marathwada region of Maharashtra, India. The linear interpolation method produces better results than a cubic, nearest neighbor, spline, shape-preserving, and PCHIP interpolation methods. The accuracy of linear interpolation methods is 90.92%. The accuracy of the cubic and PCHIP interpolation techniques is 90.76%, the accuracy of the shape-preserving and spline interpolation methods is 88.93%, and the accuracy of the nearest neighbor is 90.70%. Normal, heavy, and low rainfall are only 30–35%, 20–25%, and 35–40%, respectively. Marathwada region rainfall increased by 40 mm from 2021 to 2020. The annual temperature is 36–37°C in the Marathwada region, due to this Marathwada region falls under drought. So there is a need to change the crop pattern in the Marathwada region like temperature tolerant crops.

Least squares smoothed k-nearest neighbors online prediction of the remaining useful life of a NASA turbofan

An accurate prediction of the Remaining Useful Life (RUL) of aircraft engines plays a fundamental role in the aerospace field since it is both mission and safety critical. In fact, a reliable estimate of the RUL can effectively reduce the maintenance costs while fostering safety. This paper proposes a novel data-driven method to increase accuracy of the RUL prediction for real-time prognostic systems, considering multiple degradation mechanisms and making the model easy to implement. The proposed method exploits a novel modified k-Nearest Neighbors Interpolation (kNNI) with an a posteriori Least Square Smoothing (LSS) automatically optimized to obtain the minimum prediction error. The LSS novel formulation was also generalized and proved to be equivalent to a Cumulative and Moving Average (CMA) mixture filter, which can be easily implemented online. The method was developed and validated based on a new NASA dataset generated by the dynamic model Commercial Modular Aero-Propulsion System Simulation (N-CMAPSS) with run-to-failure data related to a small fleet of aircraft engines under realistic flight conditions. Finally, a reference kNN-based method already known in the literature was compared to the novel proposed one to demonstrate the goodness of the results and the performance improvements.

Analysis and solution to aliasing artifacts in neural waveform generation models

In recent years, with the application of deep learning in speech synthesis, waveform generation models based on generative adversarial networks have achieved high quality comparable to natural speech. In most waveform generators, a neural upsampling unit plays an essential role as it is employed to upsample acoustic features to the sample point level. However, aliasing artifacts are observed in the generated speech regardless of whether transposed convolution, subpixel convolution, or nearest neighbor interpolation are used as temporary upsampling layers. Non-ideal upsampling filters produce aliasing, according to the Shannon-Nyquist sampling theorem. This paper aims to systematically analyze how aliasing artifacts are produced in non-ideal upsampling-based waveform generators. We investigate the HiFi-GAN and VITS generation processes and discover that high-frequency spectral details are generated based on low-frequency structures using the nonlinear transformation. However, the nonlinear transformation was unable to completely remove the low-frequency spectral imprint, which eventually manifested as spectral artifacts in generated waveforms. To suppress aliasing artifacts, a low-pass filter is applied after the upsampling layer, but this results in significant performance drops. The experimental results also show that aliasing speeds up the training process by filling high-frequency vacancies. In this regard, we propose to mix high-frequency components into low-pass filtered features, allowing models to converge faster while naturally avoiding artifacts. In addition, to assess the efficacy of our method, we created an artifact-detection algorithm based on structural similarity.

A precise geoid model of Bosnia and Herzegovina by the KTH method and its validation

We compute the first gravimetric geoid model for Bosnia and Herzegovina (BHG2022) based upon the KTH method and additive corrections. The BHG2022 is computed with the help of 34,820 terrestrial gravity points and a digital elevation model produced by the Shuttle Radar Topography Mission. The gravity data is gridded onto a 0.02 arc-degree resolution via the nearest neighbour interpolation method. ITU_GGC16 and ITG-Grace2010 models are utilised to provide long wavelengths of gravity field up to 280 and 180 maximum degree/order in the geoid computation, respectively. On the basis of an external evaluation by 609 Global Navigation Satellite System (GNSS)-levelling points homogeneously distributed over the country, the accuracy of the BHG2022 is estimated to be 5.65 cm absolutely. This is the most precise geoid model ever seen in this territory. The result highlights the significant effect of all gravity data obtained from various sources on the accuracy of the geoid model.