SciELO - Scientific Electronic Library Online

 
vol.68 issue2  suppl.TIProcDETECTION OF MESOSCALE OCEANIC FEATURES USING RADARSAT-1, AVHRR AND SEAWIFS IMAGES AND THE POSSIBLE LINK WITH JACK MACKEREL (TRACHURUS MURPHYI) DISTRIBUTION IN CENTRAL CHILEFEATURING ENSO 1997-2000 IN CENTRAL CHILE author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

Share


Gayana (Concepción)

Print version ISSN 0717-652XOn-line version ISSN 0717-6538

Gayana (Concepc.) vol.68 no.2 suppl.TIProc Concepción  2004

http://dx.doi.org/10.4067/S0717-65382004000200007 

 

Gayana 68(2) supl. t.I. Proc. : 40-47, 2004 ISSN 0717-652X

USE OF SATELLITE WAVE DATA IN THE WORLDWAVES PROJECT

 

Steve Barstow, Gunnar Mørk, Lasse Lønseth & Peter Schjølberg

Fugro OCEANOR, N7462 Trondheim, Norway Email: sbarstow@oceanor.com


ABSTRACT

The demand for reliable information on wave conditions, in particular at coastal sites, is increasing. WorldWaves is a newly completed software package, developed by Fugro OCEANOR with funding from industry and within the framework of various European Commission sponsored contracts, notably EnviWave. The system simplifies the modelling of wave conditions in coastal waters, resulting in more timely, lower cost but reliable data. Thus, WorldWaves should be particularly beneficial in developing countries. In WorldWaves, the following are integrated under a single Matlab toolbox:

· High quality long-term wave data offshore all global coasts.
· Global bathymetric and coastline data.
· Two shallow water wave models: SWAN, a third generation wave model and a backward ray-tracing model.
· Sophisticated offshore and nearshore wave statistics toolboxes.
· A geographic module allowing the user to easily zoom in on geographic maps together with tools to assist the user in setting up the model grid.
· A bathymetry editing tool allowing easy editing of the bathymetric data.
· A facility allowing users to easily import their own offshore wave data

In this paper, we describe the role of satellite wave measurements in enhancing the quality of the final product with examples and applications from South America.


 

INTRODUCTION

In the absence of long-term wave data collected at a coastal site, the calculation of wave statistics requires various data sets to be assembled, including temporally and spatially long term representative directional wave data offshore of the site, bathymetric and coastline data. Further, a suitable wave model is required, capable of modelling the transfer of the offshore conditions to the site, and incorporating the relevant shallow water wave phenomena. In 2001, Fugro OCEANOR launched the Eurowaves MATLAB Toolbox (Barstow et al., 2000), a software package integrating, for all European waters, all the necessary data and tools to allow the user to quickly (in a matter of minutes) make provisional calculations of the long-term wave conditions at any coastal site along the European coastlines. It also contained all the tools to carry out more in-depth wave modelling. Eurowaves was developed under the European Commission MAST research project Eurowaves, contract no. MAST3 - CT97 ­109.

Recognising that all the data sources used in Eurowaves were also available globally, it was early recognized that a global extension to Eurowaves would be a desirable further development. In late 2001, OCEANOR were approached by a consortium of dredging companies (SSB1) in the Netherlands. SSB were interested in developing a software package which was capable of making better estimates of wave conditions (ideally, time series of wave height, period and direction) at coastal sites worldwide as input to a simulation program for the workability of dredgers, used mainly in connection with tendering work, as well as enabling estimates of extreme conditions to be made. Up to that point, offshore statistics from available on-line satellite data and offline wave atlases were the main sources consulted in the absence of in-situ wave measurements.

The new package would allow the SSB companies to evaluate the wave conditions at any coastal site worldwide using the same data sources and methodology in a quick and cost-effective manner.

In this paper, we describe the various components of WorldWaves and illustrate its capabilities with case studies for la case study in South America.

Offshore wave and wind data base

The WorldWaves package had to be capable of calculating wave conditions along any coast worldwide, i.e., coasts of North and South America (including the Gulf of Mexico), Europe (including the Mediterranean, the North Sea, the Baltic, the Black Sea and the Atlantic Islands), Africa, the near East (including the Gulf and the Caspian), the Far East (Malaysia, Indonesia, the Philippines, the South China Sea and Gulf of Thailand, Japan, China, Taiwan and Russia), Australasia and the Pacific Islands. In the construction of the WorldWaves offshore database, three types of wave data have been integrated using similar methodology to that adopted successfully in the Eurowaves project. These are a) Numerical wave model data; b) Satellite measurements; and c) In-situ measurements. The model data are used as the primary source of wave data as these data are available globally and contain all the necessary information for the offshore wave and wind input in the required time series format. The model data are quality controlled and validated and bias is corrected relative to the satellite and in-situ data.

In-situ data

In offshore waters around the world, long-term buoy wave measurement networks are still relatively few and far between. Networks with directional measurements (directional information is essential for coastal prediction) are even scarcer.

The most important networks are

a) The NOAA-NDBC buoy networks in the US (covering East and West coasts of the USA, the Gulf of Mexico and the Hawaiian Islands), in addition to the more recent Canadian network. Unfortunately, only one of the US buoys currently measures wave direction.

b) Japanese ODAS buoys in the Pacific, Sea of Japan and East China Sea.

c) The Indian National Data Buoy Programme (see http://www.niot.ernet.in/ndbp/index.html) which is currently probably the largest national program with deep ocean directional buoys used as a standard.

d) National networks in Spain, Greece, France and Italy although most buoys are rather too close to the coast to be of major value in developing our offshore wave database.

e) Long term measurements carried out in Norwegian waters and the North Sea for the offshore industry, although long term buoy data sets are mostly from measurements in the 1980s and early 1990s and are not concurrent with the model data.

This includes several Seawatch partnership programmes lead by Fugro Oceanor.

The buoy data are used primarily for a) validating satellite altimeter and wave model wave heights and b) wave model wave periods and directions. Nevertheless, in cases where buoy data do exist these are in most cases the most accurate data for the location in question and facilities are available in WorldWaves for using these data directly as the offshore input to the shallow water models. In a few cases, simultaneous measurements in deep and shallow water are available and these are very useful in validating the entire WorldWaves methodology.

Satellite data

The back-scattered signal from satellite altimeters, when calibrated and quality controlled, can provide significant wave height measurements close to the accuracy of a buoy from an orbit of typically 1,000 km. (see, for example, Krogstad and Barstow, 1999). Measurements are made each second, whilst the satellite flies over a repeat net of ground tracks at about 6 km/s. This provides enormous amounts of wave data worldwide, and with, at present, a steady flow of new data from 3 or more operational satellites, millions of new observations are becoming available each month. Global long-term satellite altimeter measurements have been performed during 1985-1989 by the US Navy's Geosat and the Geosat-Follow-on mission from 1998, by ESA's ERS-1 (from 1991 to 1996), ERS-2 (1995 - ongoing ), EnviSat (launched in 2002), and most importantly for our purposes, the US/French Topex/Poseidon mission from 1992 to 2002. Mean significant wave heights (fully calibrated; see later in this section) along all Topex/Poseidon ground around the Indian sub-continent are shown in Figure 1. The Topex Follow-on mission (Jason) was launched in 2001 and operates on the same ground track as Topex. In the meantime, Topex continues to operate on an orbit midway between the Jason orbit. Within the framework of the EC EnviWave project, Fugro OCEANOR and partners are working on validating both EnviSat and Jason satellite altimeter data.

Each satellite altimeter has to be validated in order to remove the altimeter-dependent biases on significant wave height and this is generally and most reliably done by comparing with long-term offshore buoy data, although at the beginning of missions cross-validation against other altimeters and wave model data are also used. As we are comparing temporally varying significant wave height data from buoys with spatially varying data from the satellite it is important that the buoy data are measured in areas where the gradients in significant wave height are rather small, which means in practice moorings far from coasts (ideally, several hundred kilometres) as wave conditions often vary rather strongly near to coasts both due to geographical sheltering effects, fetch limited wave growth in winds blowing offshore and shallow water effects.

We are mostly concerned here with the Topex satellite altimeter as this mission runs in parallel with the available model data ( see the next section). The other satellite data sets which could have been used are the ERS-1 and ERS-2 missions. However, these data have been routinely assimilated into the wave model and are therefore not independent of the model data. For Topex, the original altimeter (Side A) began to degrade around 1996-1997 and the bias increases slowly until the reserve altimeter was turned on in February 1999. For Side A, we use the algorithm due to Challenor and Cotton (1998) and, for Side B, we use the recent validation for the period up to June 2002 by Mørk (2003). The high accuracy is apparent in the satellite - buoy comparison shown in Figure 2.

All Topex altimeter data globally for 1992 to 2002 have been analysed applying the bias corrections above as well as an automatic data control, removing, for example, unphysical along-track variations in wave height.

Figure 1: Mean significant wave heights (1992-2002) in Chilean and Argentinian waters along all Topex/Poseidon ground tracks (from World Wave Atlas 2.0).


Figure 2: Comparison of significant wave height between the Topex Side B altimeter for 1999-2002 and 13 NOAA buoys; coincident data for Side B (from Mørk, 2002). The best fit regression line shown in the figure was used for correcting Topex Side B wave heights.

Wave model data

Nowadays, sophisticated wave models are run operationally at many meteorological centres in Europe and dedicated long-term hindcasts have also been performed. The wave models attempt to replicate the growth, decay and propagation of ocean waves based on input winds over the area in question. In practice, the limiting factor is the accuracy of these input wind fields, to which the wave results are sensitive.

In the Eurowaves project, the database from operational runs of the WAM model at the European Centre for Medium-Range Weather Forecasts (ECMWF) was selected as the best available. In the intervening years, in various studies for oil companies and coastal engineering companies, we have had the opportunity to carry out comparisons of the operational WAM data and most other major global wave model databases. In almost all cases, the WAM data have proven to out-perform the alternatives. Therefore, it was decided to choose the ECMWF database as the main component of the offshore database in WorldWaves.

Fugro OCEANOR therefore negotiated an agreement with ECMWF to initially obtain a 10-year global database for use in WorldWaves. The database consists of 6-hourly time series of significant wave height, wave period and direction for both wind sea and swell. In addition, wind speed and direction are available. The following datasets are being used:

ERA-40 data for 1994 to December 1996. ERA-40 is a global 40-year hindcast project being carried out by ECMWF (see http://www.ecmwf.int/research/era/Project/index.html)

Operational data for December 1996 to 2004 on a 0.5° lat-lon grid.

For the Mediterranean, Black Sea and the Baltic, the model was run on an 0.25° grid and data are available on a 0.5° grid for 1992 to 2002.

The WAM wave model (Komen et al., 1994) has been operational at ECMWF since 1992.

Validation and calibration of the WAM data

We briefly outline the methodology for constructing the offshore wave data base globally here.

· Initially, we have selected grid points from the WAM database globally along all coastlines at which offshore wave data is required. The selection on the 0.5° grid gives a total of some 8,000 points globally (see the points off South America in Figure 6) and includes both continental, oceanic and inland sea coastlines. In addition, for enclosed seas such as the Gulf of Mexico all grid points are included.

· The closest along-track locations of the Topex mission to each grid point are found and time series of the calibrated Topex wave heights and wind speeds are extracted. Scatter plots are then produced for the matched in time and space WAM against Topex significant wave heights. On a regional basis, any year-to-year trends in the accuracy of the WAM data have also been investigated, so that they can be corrected for if necessary.

· For some locations (NOAA buoys and a few others) we are also validating wave heights, periods (and directions), again checking for trends.

The information from the validations are then used on a point-by-point basis to correct and homogenise the wave model data.

In Figure 3, we show a map of the correlation coefficient between WAM and Topex significant wave heights globally. We can see that the correlation at most points is rather good. The bias (not shown) is also rather small, although the model underestimates a little at high sea states and overestimates in low seas (it is this bias that we correct in our calibration procedure to give the best offshore wave conditions for input to the offshore-to-nearshore wave transformation models). For South America, the results are shown in Figure 4.

Figure 3: Correlation coefficient derived from comparing Hm0 from the WAM model against Topex satellite altimeter measurements.


Figure 4: Correlation coefficient for Hm0 from the WAM model against Topex satellite altimeter measurements for South American offshore validation locations.

Further, we have compared the WAM significant wave heights, Hm0, against long term NOAA buoy data in Figure 5 for the bias and residual scatter index (standard deviation divided by the mean). In addition, comparing with Topex data rather than buoy data in exactly the same locations gives the results in Figure 6. One can see that the magnitude of both the bias and the RSI are very similar. This confirms the similar accuracy of the Topex satellite altimeter and buoy Hm0 measurements and gives us confidence of our use of Topex data globally for calibrating the model data.

During the period 1986 to 1995, the US NOAA-NDBC network included a buoy in the deep ocean off northern Chile in position 18°S, 85.1°W. We have compared the closest WAM grid point to the buoy, calibrated using the same procedure described above, and then compared the resultant significant wave height against the simultaneous buoy data. The result shows good agreement as expected (see Figure 6). The correlation coefficient is a little over 0.9. We note that the comparison with Topex gave the same correlation.

Figure 5: Validation statistics of Hm0 from WAM against NOAA data for 1996 to 2002. Left: Mean difference in cm; Right: Residual Scatter Index in %.


Figure 6: Validation statistics of Hm0 from WAM against Topex for the locations of the NOAA buoys in Figure 5. Left: Mean difference in cm; Right: Residual Scatter Index in %.

Figure 6: Comparison of the final calibrated WAM data against the NOAA-NDBC buoy which was moored to the west of northern Chile from 1986 to 1995. The comparison is more the whole of 1994.

Figure 7: The final offshore database contains about 8,000 grid points along all global coastlines. The offshore points in South America are shown. Wave data are calibrated in each point according to the results of the WAM vs. Topex validations at neighbouring point.

CONCLUSIONS

We have described the software package in some detail elsewhere (e.g., Barstow, Mørk et al., 2003). In this paper, we have described the role of satellite wave and wind data in enhancing the quality of the offshore wave data, providing the highest quality global offshore database used as input into the coastal wave models integrated in the WorldWaves software package. The full package will be demonstrated to interested parties at Porsec 2004.

ACKNOWLEDGEMENTS

The authors would like to acknowledge the members of the SSB research project "Werkbaarheid en Golfcondities" (Workability and Wave Conditions) for their financial support with respect to developing the WorldWaves software package.

We would also like to thank the European Commission for support on some aspects of this work within EnviWave (Project number EVG1-CT-2001-00051) and ARION (Project Number : IST-2000-25289).

1SSB (Stichting Speurwerk Baggertechniek) is an organisation carrying out joint research and development for the 3 major Dutch dredging companies Van Oord, Boskalis and Ballast Ham.

REFERENCES

Barstow, S.F.; Athanassoulis, M. & L. Cavaleri (2000) "EUROWAVES: Integration of data from many sources in a user-friendly software package for calculation of wave statistics in European coastal waters." Proc. Oceanology International 2000 Conference, Brighton, UK, March 2000, pp. 269-277 (CD-ROM). [         [ Links ]1]

Barstow, S.F., G. Mørk, L. Lønseth, P. Schjølberg, G. Athanassoulis, K. Belibassakis, Th. Gerostathis & G. Spaan "WorldWaves: High quality coastal and offshore wave data within minutes for any coastal site", Proceedings of the COPEDEC 2003 confernce, Colombo, Sri Lanka, September 2003. [         [ Links ]2]

Challenor, P. & D. Cotton (1998) "Trends in TOPEX significant wave height measurement. Paper presented at the TOPEX/Poseidon Working Team meeting, Keystone, Colorado, USA, October 1998. [         [ Links ]3]

Komen, G.J., Cavaleri, L., Donelan, M., Hasselmann,K., Hasselmann, S. & P.A.E.M. Janssen (1994) "Dynamics and Modelling of Ocean Waves", Cambridge University Press, 532 pp. [         [ Links ]4]

Krogstad, H.E. & S.F. Barstow (1999) "Satellite Wave Measurements for coastal engineering applications." Coastal Engineering, 37, 283-307. [         [ Links ]5]

Mørk, G. (2003) "Validation of the Topex Side B altimeter with respect to significant wave height and wind speed". OCEANOR Report No. OCN R-22022. [         [ Links ]6]

 

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License