Page 188 - NGTU_paper_withoutVideo

Modern Geomatics Technologies and Applications

A New Approach of Waveform Re-Tracking for Monitoring Sea Surface Topography in
the Strait of Hormuz

1*
1
Arash Amini , Parisa Agar , Majid Mostafavi , Ali Sabilian , Shirzad Roohi 4
2
3

1 Faculty of Geodesy and Geomatics Engineering, K. N. Toosi University of Technology, Tehran, Iran
2 Department of Civil Engineering and Architecture, Tallinn University of Technology, Tallinn, Estonia
3 Faculty of Technical and Engineering, North Tehran Branch, Azad University, Tehran, Iran
4 Department of Surveying Engineer, South Tehran Branch, Azad University, Tehran, Iran
* parisaagar1995@email.kntu.ac.ir

Abstract: Satellite altimetry techniques has many challenges over inland water bodies and coastal area due to none-water
reflected pulses which cause waveform corruption. Consequently, standard waveform shape which have been designed
for open oceans monitoring are no longer valid in these areas. This makes specially designed new re-tracking algorithm
development essential delegated for these regions. In this study apart from available re-trackers in Level-2 data, three
other re-tracking approaches including: original waveform, first meaningful sub-waveform and mean waveform per each
pass and cycle have compared with the new developed method called maximum correlation with mean waveform. Sentinel-
3 (SRAL) altimetry data over Strait of Hormuz located in Persian Gulf during 2016/06/07 to 2019/06/11 examined with
the available Tide Gauge (TG) station data record in study area. Analysis of distorted waveform showed that three passes
(154, 595, 709) have the highest corruption rates compare to the other available passes in this area (45.92%, 38.6% and
30.97%). Consequently they have been selected for further steps of the study. Sea Surface Topography (SST) time series
estimated from with proposed re-tracking approach in terms of RMSE with respect to TG records. Our assessments
showed 34 and 29 cm improvement in RMSE for pass 599 and 709 comparing to level-2 data. Overall time series also
showed 17 cm improvement in RMSE of SST derived from the proposed approach in compare with level-2 data.

Keywords: Satellite Altimetry, Coastal Area, Waveform Re-tracking, Threshold, Maximum Correlation

1. Introduction
Global mean sea level rise due to climate changes could lead to future natural hazards including flooding, shoreline
changes consequently, coastal erosion and salinization of estuaries and of coastal aquifers. Populated cities are located in costal
zones are extremely exposed to a variety risks of these hazards. Therefore, the sea level changes monitoring crucial at these areas
[1]. Tide gauge stations provide valuable and independent information about sea level monitoring with some limitations such as
differences in vertical datums, spatial and temporal resolutions. These observations also include coastal effects such as land
movements or wave and wind surges. Therefore, the complexity of coastal sea level is difficult to represent with only sparse tide
gauges along the coast. Satellite altimetry (SA) could overcome some of the constraints by providing an accurate, unique, and
consistent data for monitoring sea levels from the open ocean up to the coasts[2-4]. SA estimate the altimeter range by measuring
two ways travel time of the emitted electromagnetic pulse to the surface. A part of these times is related to interaction of the
pulse with the surface in term of returned power of the waveform. This time divide to specific intervals called gate, so the
waveform could be defined as received power to the radar according to the gate [4]. Over open ocean received waveforms follow
a standard shape form with steeply rising leading edge followed by a trailing edge with gradually diminishing power. This
standard shape agrees with the theoretical Brown model [5]. In this standard model the corresponding gate with middle point in
leading edge, which is the default altimeter waveform, shows the time of received echo form the surface. But altimetry suffers
the distortion in waveforms due to land contaminated at altimeter footprints over coastal area (~30km offshore). In this areas the
altimeter’s pregiven gate, which has been defined to be used for open ocean, would lead to incorrect surface estimation. This
distorted waveform error called the re-tracking error [6]. Waveform re-tracking algorithms could be based on analytical (curve-
fitting algorithm) or empirical (statistics of the waveform) models [7]. The β -parametric fitting algorithm was firstly presented
by Martin et al in 1983 to estimate the SeaSat altimeter range to the ice-sheets[8]. Wingham et al. put forward the OCOG (Off-
Centre of Gravity) algorithm method based on the statistical characteristics of waveforms to define centre of gravity of any

1