The reliability of vertical differences in temperature in the ocean is essential for the accuracy of this assessment. In general, oceanic water masses acquire their temperatures and salinities through heat and water exchange with the atmosphere. Once they are no longer in contact with the atmosphere, their temperature and salinity vary only very slowly. Waters that are found between 700 to 1200 m depths in the tropical ocean were last in contact with the atmosphere in the higher mid-latitudes of the southern Ocean, between roughly 50 and 60° of latitude. Waters in the upper 100 m of the tropical ocean are still influenced by the atmosphere and are subject to seasonal variations.
Whereas temperatures in the open ocean can vary between 2 and 28°C, salinities vary only from about 34 to 37 psu (practical salinity units), with higher (resp. lower) values being found in enclosed evaporation (resp. precipitation) basins, such as the Mediterranean Sea or the Red Sea (resp. the Baltic Sea). However, the effect on the density of seawater of a change in salinity of 1 psu is generally equivalent to a temperature change of 4°C. The temperatures and salinities used in this assessment are monthly mean values computed from the World Ocean Atlas and the MyOcean project.
World Ocean Atlas
The World Ocean Atlas 2009 (WOA09) is a set of so-called objectively analyzed climatological fields of in-situ measurements of temperature, salinity, dissolved oxygen, Apparent Oxygen Utilization (AOU), percent oxygen saturation, phosphate, silicate, and nitrate. As of the 2009 update, it includes more than 9,000,000 temperature profiles collected from ships or from floats throughout the world.
In this objective analysis, a number of vertical levels (33 for WOA) and a 1-degree global grid are defined. For each grid cell, neighbouring measurements are interpolated and weighed according to a defined influence radius and distance to the grid cell in order to compute a value representative for this 1 x 1-degree cell. In addition, some corrections are also applied if the available measurements provide sufficient information.
The fields are available as climatological monthly means (also seasonal and annual) for the following decades: 1955-1964, 1965-1974, 1975-1984, 1985-1994, and 1995-2006. The word climatological refers to the fact that the values provided by the product for January (for instance) are computed as the average of all months of January for the considered decade. The information presented in the Ocean Potential tool uses monthly mean temperature and salinity profiles for the decade 1995-2006.
The following version of the World Ocean Atlas (World Ocean Atlas 2013, WOA13) is currently being released and will be used within the Ocean Potential tool as a replacement to WOA09. Although the horizontal resolution has been increased to 1/4°, it is built following the same principles as WOA09.
Data and data description are available at:
- Locarnini, R. A., A. V. Mishonov, J. I. Antonov, T. P. Boyer, H. E. Garcia, O. K. Baranova, M. M. Zweng, and D. R. Johnson, 2010. World Ocean Atlas 2009, Volume 1: Temperature. S. Levitus, Ed. NOAA Atlas NESDIS 68, U.S. Government Printing Office, Washington, D.C., 184 pp. Available online at http://www.nodc.noaa.gov/OC5/indprod.html
- Antonov, J.I., D. Seidov, T.P. Boyer, R.A. Locarnini, A.V. Mishonov, H.E. Garcia, O.K. Baranova, M.M. Zweng, and D.R. Johnson, 2010. World Ocean Atlas 2009 Volume 2: Salinity. S. Levitus Ed. NOAA Atlas NESDIS 69, U.S. Gov. Printing Office, Washington, D.C., 184 pp. Available online at http://www.nodc.noaa.gov/OC5/indprod.html
MyOcean model results
MyOcean is a series of projects granted by the European Commission within the GMES Program (7th Framework Program for European Research and Development), whose objective is to define and to set up a concerted and integrated pan-European capacity for ocean monitoring and forecasting. MyOcean provides state-of-the-art information available on the Global Ocean (worldwide coverage) and on European seas, based on the combination of space and in-situ observations, and their assimilation into numerical models.
In particular, the Ocean Potential tool relies on model output from the Global Ocean Physics Reanalyses GLOBAL_REANALYSIS_PHYS_001_009 and 010.
A reanalysis is a scientific method through which observations and a numerical model that simulates one or more aspects of the Earth system are combined objectively to generate a synthesized estimate of the state of the system. A numerical model of the global ocean is a numerical representation of the ocean. The world is divided into a number of points in the three spatial dimensions. At each point, the driving equations are discretized and solved iteratively.
A reanalysis is an attempt at simulating the (recent) past climate and/or ocean state over several decades as accurately as possible, by objectively combining past observations and a numerical model through data assimilation techniques. Data assimilation proceeds by analysis cycles. In each analysis cycle, observations of the current (and possibly past) state of a system are combined with the results from the numerical model (the forecast) to produce an analysis, which is considered as ‘the best’ estimate of the current state of the system. This is called the analysis step. Essentially, the analysis step tries to balance the uncertainty in the data and in the forecast. The model is then advanced in time and its result becomes the forecast in the next analysis cycle.
The Global Ocean Physics Reanalysis GLOBAL_REANALYSIS_PHYS_001_009 is based on the GLORYS2V1 global reanalysis of Mercator Ocean for the period 1993-2009. This reanalysis produced monthly means of temperature, salinity, ocean currents, sea surface height and sea ice parameters, at ¼-degree horizontal resolution with 75 vertical levels. The atmospheric conditions originate from the ERA-Interim atmospheric reanalysis. The ERA-Interim reanalysis is a reanalysis of the atmospheric state from the year 1979 to date. The reanalysis for the ocean state in GLORYS2V1 uses assimilation of (in-situ) vertical temperature and salinity profiles, and observations of sea surface temperature and sea level anomalies from satellite.
The Global Ocean Physics Reanalysis GLOBAL_REANALYSIS_PHYS_001_010 relies on the UR025.4 reanalysis of the University of Reading for the period 1993-2009. This reanalysis also produced monthly means of temperature, salinity, ocean currents, sea surface height and sea ice parameters, at ¼-degree horizontal resolution with 75 vertical levels. The atmospheric conditions originate from the ERA-Interim atmospheric reanalysis. However, the data assimilation method differs from that used in the GLORYS2V1 system. The reanalysis for the ocean state uses assimilation of in situ and satellite sea surface temperature, sea level anomalies, in-situ vertical temperature and salinity profiles, and satellite sea-ice concentration.
Data and data description are available at:
Performance and limitations
Information on the vertical structure of the ocean temperature is essential to the pre-assessment. Even though the use of drifting profiling floats in the recent years has allowed to finally sample the whole global ocean, the density of measurements (in both time and space) does not yet allow to diagnose the vertical structure of a specific region at a scale finer than about one degree in latitude and longitude. In this respect, numerical models present the advantage of covering the whole world with finer horizontal and vertical resolutions.
Because of the way they are built, reanalyses can be considered as model-assisted interpolations in space and time of observations on a global scale for a given period. The model is constrained into computing a solution that physically satisfies all the underlying equations (thus a physically acceptable solution), but also into reproducing the observed state as closely as possible given these limits, and some defined uncertainties in the observations. However, it should be kept in mind that a reanalysis should not be equated with either “observations” or “reality”. Some variables respond more or less well to the observational constraints, so that the reliability of the reanalysis may vary depending on the location, time period, and the variable considered. In addition, the changing mix of observations, and certain biases in observations and models, may result in spurious variability and trends.
However, it should also be noted that the MyOcean products are still aimed at reproducing the large-scale oceanic patterns, in which tides play only a marginal role. As a result, most global ocean models (including the ones the MyOcean products rely on) do not include tides. Only in shallow coastal seas may the effect of tides on current patterns and temperature become larger than that of winds and density-driven currents. However, this only applies when the depth becomes small enough (order a few tens of meters).
Compared to other state-of-the-art numerical models, the horizontal resolution of the global MyOcean products (1/4°) remains relatively limited and some finer models exist. For instance, the HYCOM model is currently being used for a global forecast of the world ocean with an horizontal resolution of 1/12°, which could have provided a suitable basis for the pre-assessment. However, this HYCOM experiment focuses on short-term forecasts of the ocean state, whereas the selected MyOcean products are reanalyses (and thus a recomputation of the past ocean state as closely as possible). In a forecast such as that with the HYCOM model, past observations are used to constrain the model up until the moment from which the forecast begins. The forecast is thus computed without any knowledge of the future (observed) state. As a result, the model simulation may be more sensitive to the so-called model drift, a tendency of the model to drift away from the observed state and exhibit a systematic bias. In a reanalysis, the model is constrained at each step by the past and the future observations, so that the model drift is more limited. For the quality of this assessment it was decided that the model results should be as close as possible to the real, observed ocean state, hence the choice for MyOcean products.
The fact remains that the products used for the assessment tool are at the state-of-the-art of ocean modelling and forecasting. Yet, it should be pointed out that these products are only numerical representations of an extremely complex system, and have therefore limitations. Some originate from the inherent limitations to the discretization and resolution of the underlying equations, others from the quality of the atmospheric forcing, from the current understanding of certain relevant processes and the way their effect is parameterized into the model. In addition, the numerical models used cannot resolve any details smaller than the size of an elementary grid cell 1/4° or about 45 km at the Equator).
The assessment focuses on providing average temperature patterns in the vicinity of a given location, which makes its results less likely affected by inaccuracies and product limitations. Temperatures and salinities in the deep ocean are primarily driven by large-scale mechanisms in the world ocean and are therefore fairly insensitive to very small-scale features. Although temperatures and salinities near the surface may be more strongly influenced by smaller-scale features not captured by the global models, the mean values should be correct.
A specific assessment of both products indicates that the global mean sea surface temperature in GLOBAL_REANALYSIS_PHYS_001_009 differs with less than 0.5°C RMS compared to measurements. GLOBAL_REANALYSIS_PHYS_001_010 also compares well to the global mean sea surface temperature. With this latter product, temperatures in the upper ocean are higher than those of climatological observation-based atlases (World Ocean Atlas WOA09) but it is actually thought that this corresponds to the recent observed warming of the ocean that the World Ocean Atlas does not capture properly because of the presence of old data (Dombrowsky et al. 2013).
Sea surface salinities in the tropical regions of GLOBAL_REANALYSIS_PHYS_001_009 were found to be about 0.2 to 0.8 psu too low compared to observations. The picture is less uniform in GLOBAL_REANALYSIS_PHYS_001_010 with some areas being saltier and others fresher. In general, it is believed that both models perform less well near the surface because an overestimate of the precipitation flux in the atmospheric product used for forcing the model. Below the surface, the agreement with measurements for both products is better.
In general, the Ocean Potential tool should therefore provide a reliable indication of the ocean temperatures that can be expected in a given area within about 0.5°C, and thus from its general suitability for installing an OTEC system. However, for a more detailed evaluation of certain design, cost and profitability aspects it is advisable that a more detailed site-specific study be undertaken.
Dombrowsky E., N. Ferry, L. Parent, S. Masina, A. Storto, K. Haines, M. Valdivieso, B. Barnier, J.M. Molines, 2013. Quality Information Document For Global Ocean Reanalysis Products GLOBAL-REANALYSIS-PHYS-001-004 GLOBAL-REANALYSIS-PHYS-001-009 GLOBAL-REANALYSIS-PHYS-001-010 GLOBAL-REANALYSIS-PHYS-001-011. MyOcean.eu report MYO2-GLO-PUM-001-004-009-010-011, Issue 2.3, 14 Feb. 2013.
Dee D., J. Fasullo, D. Shea, J. Walsh, & National Center for Atmospheric Research Staff (Eds). Last modified 06 Sep 2013. The Climate Data Guide: Atmospheric Reanalysis: Overview & Comparison Tables. Retrieved from https://climatedataguide.ucar.edu/climate-data/atmospheric-reanalysis-overview-comparison-tables.