Prediction of geoid undulations: Random forest versus classic interpolation techniques

Akar A., Konakoğlu B., Akar Ö.

CONCURRENCY COMPUTATION PRACTICE AND EXPERIENCE, cilt.34, sa.18, ss.1-11, 2022 (SCI-Expanded)

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 34 Sayı: 18
  • Basım Tarihi: 2022
  • Doi Numarası: 10.1002/cpe.7004
  • Dergi Adı: CONCURRENCY COMPUTATION PRACTICE AND EXPERIENCE
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Aerospace Database, Applied Science & Technology Source, Communication Abstracts, Compendex, Computer & Applied Sciences, INSPEC, Metadex, zbMATH, Civil Engineering Abstracts
  • Sayfa Sayıları: ss.1-11
  • Erzincan Binali Yıldırım Üniversitesi Adresli: Evet

Özet

Local geoid determination studies are commonly carried out today to establish therelationship between the ellipsoidal height(h)determined by satellite geodesy meth-ods and the orthometric height(H)found using geoid undulation(N). The aim of thisstudywastodeterminealocalgeoidusingthekriging,localpolynomial(LP),andinversedistance to a power (IDP) interpolation methods along with the random forest (RF)regression method and to compare the performance. For the application, 193 GlobalNavigation Satellite System (GNSS)/leveling points homogeneously distributed overthe study area were selected as reference data, and 70 GNSS/leveling points wereselectedastestdata.Theresultswerecomparedintermsofaccuracyusingwell-knownperformancemetrics,namely,therootmeansquareerror(RMSE),meanabsoluteerror(MAE), and coefficient of determination (R2) values. According to the results, the high-estR2and the lowest RMSE and MAE values were obtained with the RF regressionmethod. These findings demonstrated the superiority of the RF regression over theclassic interpolation methods applied in the local geoid determination for the studyarea