Satellite Wind Speed Research Articles

Assessing satellite-derived wind speeds: insights from buoy measurements and analysis of deviation factors

ABSTRACT To evaluate the latest generation of satellite wind speed from AMSR2, ASCAT, GMI, SSMIS, SMAP, and WindSat, buoy measurements were obtained from buoys in coastal and tropical open ocean areas. Root mean square errors were calculated, resulting in values of 1.19 m/s for AMSR2, 1.25 m/s for ASCAT, 1.37 m/s for GMI, 1.38 m/s for SSMIS, 2.48 m/s for SMAP, and 1.18 m/s for WindSat. Numerous data pairs fall above the 45-degree line, mainly due to decreased buoy ac-curacy during high winds. Factors contributing to this include underestimation of buoy readings due to rough sea states and wind profile distortion. Satellite biases exhibit unimodal fluctuations over a one-year cycle, aligning with annual sea surface temperature variations. Statistical effects in extreme wind speed range cause abnormal fluctuations in statistics. Higher latitudes experi-ence larger deviations due to rough sea state. Ocean currents affect satellite wind speed observa-tions, particularly when aligned with wind direction. Coastal wind speed gradients primarily influenced from land contamination like rugged mountainous terrain along the West Coast of North America. Coastal upwelling also contributes to bias, with the west coast showing higher bias. A global comparison between SMAP and WindSat retrieval wind speeds demonstrates SMAP’s poor ability to monitor general wind speeds and validates the satellite-buoy experiment.

Developing a new wind dataset by blending satellite data and WRF model wind predictions

This paper presents an approach to improve wind datasets developed using the regional atmospheric model Weather Research Forecasting by combining its predictions with remotely sensed wind observations in enhanced wind speed analyses that leads to blended winds. In this study, satellite data derived from scatterometers, radiometers, and synthetic aperture radar are used. The spatial and temporal features of each wind product are thoroughly analysed. For the probabilistic evaluation of their skill, comprehensive comparisons with available buoy data are carried out. The statistical analysis shows that the combined use of satellite and numerical weather prediction model data improves the agreement with buoy measurements, demonstrating the added value of using the blended product. As an application of the method, new improved satellite wind speeds are presented in the form of a wind energy assessment along the Iberian coastal area. From inspection of the provided wind power maps, northern and central regions emerge as the most promising areas for wind harnessing offshore despite some seasonal variations. Finally, potential wind farm sites are provided, along with insights into multi-year wind speed distribution. The results show how the new dataset can be used for the selection of promising areas for wind exploitation.

Effects of <sup>238</sup>U variability and physical transport on water column <sup>234</sup>Th downward fluxes in the coastal upwelling system off Peru

Abstract. The eastern boundary region of the southeastern Pacific Ocean hosts one of the world's most dynamic and productive upwelling systems with an associated oxygen minimum zone (OMZ). The variability in downward export fluxes in this region, with strongly varying surface productivity, upwelling intensities and water column oxygen content, is however poorly understood. Thorium-234 (234Th) is a powerful tracer to study the dynamics of export fluxes of carbon and other elements, yet intense advection and diffusion in nearshore environments impact the assessment of depth-integrated 234Th fluxes when not properly evaluated. Here we use vessel-mounted acoustic Doppler current profiler (VmADCP) current velocities, satellite wind speed and in situ microstructure measurements to determine the magnitude of advective and diffusive fluxes over the entire 234Th flux budget at 25 stations from 11 to 16∘ S in the Peruvian OMZ. Contrary to findings along the GEOTRACES P16 eastern section, our results showed that weak surface wind speed during our cruises induced low upwelling rates and minimal upwelled 234Th fluxes, whereas vertical diffusive 234Th fluxes were important only at a few shallow shelf stations. Horizontal advective and diffusive 234Th fluxes were negligible because of small alongshore 234Th gradients. Our data indicated a poor correlation between seawater 238U activity and salinity. Assuming a linear relationship between the two would lead to significant underestimations of the total 234Th flux by up to 40 % in our study. Proper evaluation of both physical transport and variability in 238U activity is thus crucial in coastal 234Th flux studies. Finally, we showed large temporal variations on 234Th residence times across the Peruvian upwelling zone and cautioned future carbon export studies to take these temporal variabilities into consideration while evaluating carbon export efficiency.

Comparative study of offshore winds and wind energy production derived from multiple scatterometers and met buoys

Wind resource assessment is a challenging task where the availability of offshore Met buoys is limited. Satellite remote sensing provides the best alternative inexpensive data and fills the data gaps by providing a huge volume of data for extended periods. However, due to the availability of multiple satellite scatterometers, the assessment of such resources from any single scatterometer would lead to inconsistency. The present article provides an extensive comparison of four scatterometers namely QuikSCAT, OSCAT, ASCAT-A and ASCAT-B with long-term in-situ wind datasets over the North Indian Ocean. Results show that the QuikSCAT and the OSCAT wind data have a lesser bias with the range of 0.15 m/s (2.4%) to 0.83 m/s (15.1%) before adjustments. For adjusting the satellite wind speeds, linear regression was used. Further, the synergetic approach of linear regression adjustments and the combination of scatterometer data have resulted in smaller differences. The absolute deviation in mean wind speed between co-located buoy data and combined scatterometer data is less of the order 2.1%, and the mean power density showed a difference of 2.5%. The study revealed that the combined scatterometer datasets would help in the accurate assessments of wind resource rather than a single scatterometer.

Comparison of Wind Speeds from Spaceborne Microwave Radiometers with In Situ Observations and ECMWF Data over the Global Ocean

This study compares wind speeds derived from five satellite microwave radiometers with those directly observed by buoy-mounted anemometers and the global analyses produced by the European Center for Medium-Range Weather Forecasts (ECMWF) model. Buoy comparisons yield wind speed root mean square errors of 0.82 m/s for WindSat, 1.45 m/s for SSMIS F16, 1.39 m/s for SSMIS F17, 1.43 m/s for AMSR-E, and 1.45 m/s for AMSR2. The overall mean bias for each satellite is typically <0.25 m/s when averaged over all selected buoys for a given study time. The satellite wind speeds are underestimated with respect to the buoy observations at a band of the tropical Pacific Ocean from −8°S to 4°N. The mean buoy–satellite difference as a function of year is always <0.4 m/s, except for SSMIS F16. The selected satellite wind speeds show an obvious seasonal characteristic at high latitudes. In comparison with the ECMWF data, some obviously positive differences exist at high southern latitudes in January and at high northern latitudes in July.

Assessment of the Offshore Wind Speed Distributions at Selected Stations in the South-West Coast, Nigeria

This paper assessed the offshore wind speed distributions at 10 m height in the southwest coast of Nigeria using a high-resolution satellite observations at 0.25˚ spatial grid resolution. Satellite wind speed and direction recorded over a 10-year period (2002–2011) were derived from the CCMP L3.0 data. The monthly, seasonal and annual wind characteristics for energy conversion based on three probability distributions were assessed. The Rician and Weibull models gave better fitting of the offshore wind speed at the southwest coast compared to the Rayleigh model. Results also revealed that the monthly mean wind speed variation exhibits a rather non-monotonic trend across different grid points driven by changes in the weather system activities while further investigation reveals higher wind speed and power density at the coastal region than over land in the northern Nigeria. The monthly mean wind speed recorded at the coastal region ranges between 4.99 and 5.56 m/s, 5.32 m/s for the interannual mean wind speed with the summer (JJA) and autumn (SON) mean wind speed distributions ranging from 6.11–6.76 m/s and 4.86–6.17 m/s, respectively, at 10 m height asl. The coastal wind speed distributions show that energy conversion at the southwest region of Nigeria is viable.

Tractable Analytic Expressions for the Wind Speed Probability Density Functions Using Expansions of Orthogonal Polynomials

AbstractThe use of the two-parameter Weibull function as an estimator of the wind speed probability density function (PDF) is known to be problematic when a high accuracy of fit is required, such as in the computation of the wind power density function. Various types of nonparametric kernels can provide excellent fits to wind speed histograms but cannot provide tractable analytical expressions. Analytic expressions for the wind speed PDF are needed for many applications, particularly in the downscaling of model or satellite wind speed estimates to the regional or point scale. It is demonstrated that the judicious use of an expansion of orthogonal polynomials can produce more accurate estimates of the wind speed PDF than relatively simply parametric functions, such as the commonly used Weibull function. This study examines four such expansions applied to two different surface wind speed datasets in Oklahoma. The results indicate that the accuracy of fit of a given expansion is strongly related to how close the basis weight function in an expansion resembles the wind speed histogram. It is shown that this basis function, which is the first term in the expansion, acts as a first “best guess” to the true wind speed PDF and that the additional terms act to “adjust” the fit to converge on the true density function. The results indicate that appropriately chosen orthogonal polynomials can provide an excellent fit and are quite tractable.

Application of ALOS/PALSAR to estimate carbon dioxide transfer velocities compared with satellite wind speed data 2007–2008

Gas exchange contributes to the mitigation of an anthropogenic greenhouse effect through absorption of excess atmospheric CO2 by the oceans. The gas transfer velocity requires computation in order to describe gas exchange and interaction. The absolute calibration of the relationship between air and sea CO2 transfer velocity and wind speed (U 10) has been under debate for a long time because the global averages of the CO2 exchange coefficients, deduced from many experimental data relationships, disagreed with each other. In this research, the CO2 transfer velocity () was derived from satellite wind speeds that were computed using three different parameterizations (k–U) and was then compared with the derived from Phased Array type L-band Synthetic Aperture Radar (PALSAR) (k P). To obtain the transfer velocity of CO2 between the atmosphere and ocean using PALSAR in the Bali Sea area close to the Indian Ocean, daily sea surface temperature, wind friction velocity (U*) from PALSAR images and oceanography data were collected and analysed. Additionally, values were computed from the QuikSCAT satellite wind speed and compared to three different k–U relationships suggested by different authors. The three algorithms of parameterization were used with QuikSCAT daily data and produced a value that was close to that of derived from PALSAR (k P).

Parameterizations and Algorithms for Oceanic Whitecap Coverage

Abstract Shipboard measurements of fractional whitecap coverage W and wind speed at 10-m height, obtained during the 2006 Marine Aerosol Production (MAP) campaign, have been combined with ECMWF wave model and Quick Scatterometer (QuikSCAT) satellite wind speed data for assessment of existing W parameterizations. The wind history trend found in an earlier study of the MAP data could be associated with wave development on whitecapping, as previously postulated. Whitecapping was shown to be mainly wind driven; for high wind speeds (>9 m s−1), a minor reduction in the scatter of in situ W data points could be achieved by including sea state conditions or by using parameters related to wave breaking. The W values were slightly larger for decreasing wind/developed waves than for increasing wind/developing waves, whereas cross-swell conditions (deflection angle between wind and swell waves between ±45° and ±135°) appeared to dampen whitecapping. Tabulated curve fitting results of the different parameterizations show that the errors that could not be attributed to the propagation of the standard error in U10 remained largely unexplained. It is possible that the counteracting effects of wave development and cross swell undermine the performance of the simple parameterizations in this study.

Impact of Saharan Dust on Ocean Surface Wind Speed Derived by Microwave Satellite Sensors

In the present paper ground truth and remotely sensed datasets were used for the investigation and quantification of the impact of Saharan dust on microwave propagation, the verification of theoretical results, and the validation of wind speeds determined by satellite microwave sensors. The influence of atmospheric dust was verified in two different study areas by investigations of single dust storms, wind statistics, wind speed scatter plots divided by the strength of Saharan dust storms, and wind speed differences in dependence of microwave frequencies and dust component of aerosol optical depth. An increase of the deviations of satellite wind speeds to ground truth wind speeds with higher microwave frequencies, with stronger dust storms, and with higher amount of coarse dust aerosols in coastal regions was obtained. Strong Saharan dust storms in coastal areas caused mean relative errors in the determination of wind speed by satellite microwave sensors of 16.3% at 10.7 GHz and of 20.3% at 37 GHz. The mean relative errors were smaller in the open sea area with 3.7% at 10.7 GHz and with 11.9% at 37 GHz.

Global average of air‐sea CO2 transfer velocity from QuikSCAT scatterometer wind speeds

The absolute calibration of the relationship between air‐sea CO2 transfer velocity, k, and wind speed, U, has been a topic of debate for some time, because k global average, 〈k〉, as deduced from Geochemical Ocean Sections Study oceanic 14C inventory has differed from that deduced from experimental k‐U relationships. Recently, new oceanic 14C inventories and inversions have lead to a lower 〈k〉. In addition, new measurements performed at sea in high–wind speed conditions have led to new k‐U relationship. Meanwhile, quality and sampling of satellite wind speeds has greatly improved. The QuikSCAT scatterometer has provided high‐quality wind speeds for more than 7 years. This allows us to estimate the global distributions of k computed using k‐U relationships and temperature‐dependent Schmidt numbers from 1999 to 2006. Given the difficulty of measuring in situ wind speed very accurately, we performed a sensitivity study of the 〈k〉 uncertainty which results from QuikSCAT U uncertainties. New QuikSCAT‐buoy U comparisons in the northern Atlantic Ocean and in the Southern Ocean confirm the excellent precision of QuikSCAT U (RMS difference of about 1 m s−1), but it is possible that QuikSCAT overestimates wind speeds by 5%, leading to a possible overestimation of k derived with quadratic relationships by 10%. The 〈k〉 values obtained with two recent experimental k‐U relationships are very close, between 15.9 and 17.9 cm h−1, and within the error bar of k average deduced from the new oceanic 14C inventory.

Improved near real time surface wind resolution over the Mediterranean Sea

Abstract. Several scientific programs, including the Mediterranean Forecasting System Toward Environmental Predictions (MFSTEP project), request high space and time resolutions of surface wind speed and direction. The purpose of this paper is to focus on surface wind improvements over the global Mediterranean Sea, based on the blending near real time remotely sensed wind observations and ECMWF wind analysis. Ocean surface wind observations are retrieved from QuikSCAT scatterometer and from SSM/I radiometers available at near real time at Météo-France. Using synchronous satellite data, the number of remotely sensed data available for each analysis epoch (00:00 h; 06:00 h; 12:00 h; 18:00 h) is not uniformly distributed as a function of space and time. On average two satellite wind observations are available for each analysis time period. The analysis is performed by optimum interpolation (OI) based on the kriging approach. The needed covariance matrixes are estimated from the satellite wind speed, zonal and meridional component observations. The quality of the 6-hourly resulting blended wind fields on 0.25° grid are investigated trough comparisons with the remotely sensed observations as well as with moored buoy wind averaged wind estimates. The blended wind data and remotely wind observations, occurring within 3 h and 0.25° from the analysis estimates, compare well over the global basin as well as over the sub-basins. The correlation coefficients exceed 0.95 while the rms difference values are less than 0.30 m/s. Using measurements from moored buoys, the high-resolution wind fields are found to have similar accuracy as satellite wind retrievals. Blended wind estimates exhibit better comparisons with buoy moored in open sea than near shore.

Variability of the net air–sea CO2 flux inferred from shipboard and satellite measurements in the Southern Ocean south of Tasmania and New Zealand

We determine the distribution of oceanic CO2 partial pressure (pCO2) with respect to remotely sensed parameters (sea surface temperature (SST) and chlorophyll (Chl)) in order to gain an understanding of the small‐scale (10–100 km) pCO2 variability and to estimate the net air–sea CO2 flux in the region (125°E–205°E; 45°S–60°S), which represents 22% of the Southern Ocean area between 45°S and 60°S. We split the study area into several biogeochemical provinces. In chlorophyll‐poor regions, pCO2 is negatively correlated with SST, indicating that pCO2 is mostly controlled by mixing processes. For Chl > 0.37 mg m−3, pCO2 is negatively correlated with Chl, indicating that pCO2 variability is mostly controlled by carbon fixation by biological activity. We deduce fields of pCO2 and of air–sea CO2 fluxes from satellite parameters using pCO2‐SST, pCO2‐chlorophyll relationships and air–sea gas exchange coefficient, K, from satellite wind speed. We estimate an oceanic CO2 sink from December 1997 to December 1998 of −0.08 GtC yr−1 with an error of 0.03 GtC yr−1. This sink is approximately 38% smaller than that computed from the Takahashi et al. (2002) climatological distribution of ΔpCO2 for the 1995 year but with the same K (−0.13 GtC yr−1). When we correct ocean pCO2 for the interannual variability between 1995 and 1998, the difference is even larger, and we cannot reconcile both estimates in February–March and from June to November. This strengthens the need of new in situ measurements for validating extrapolation methods and for improving knowledge of interannual pCO2 variability.

An assessment of the effect of sea surface surfactant on global atmosphere‐ocean CO2 flux

We assess the possible impact of the distribution of naturally occurring surfactants on the direct integration of the global atmosphere‐ocean CO2 flux across the ocean surface. The global atmosphere‐ocean CO2 flux is calculated using the monthly mean ΔpCO2 climatology compiled by Takahashi et al. [1997] as well as satellite wind speed and sea‐surface temperature data. In the absence of any global map of surfactant coverage and as it is known that phytoplankton exudates and degradation products are the major sources of marine surfactants, ocean primary productivity, which can be derived from the satellite‐based estimate of chlorophyll concentration, is used as an indicator of the presence of surfactants as proposed by Asher [1997]. From the calculated results it is found that suppression of the upward and downward CO2 fluxes by marine surfactants exhibits an asymmetric effect: The average percent reduction of absorption flux by surfactants is about twice that of outgassing, which results in an overall decrease in the net global CO2 uptake by the oceans. For almost half of the year (between January and May) the presence of surfactants does not affect CO2 outgassing from global oceans. In contrast, throughout the entire year the presence of surfactants suppresses CO2 absorption by the oceans. The major reduction in absorption fluxes occurs in the northern Pacific and Atlantic (10°N to 70°N) in all seasons and in the Southern Ocean (south of 40°S) in austral spring and summer. However, the most significant decrease in outgassing fluxes occurs in the equatorial and southern Pacific Ocean (40°S to 10°N), particularly in the eastern equatorial and subtropical waters off the southern American coast, in the period of austral spring and summer. Annual net CO2 flux is reduced by approximately 20% under the surfactant coverage condition that the primary productivity is above a threshold value of 25 g‐C m−2 mom−1 and by about 50% with a threshold of 15 g‐C m−2 mom−1.

Influence of gas exchange coefficient parameterisation on seasonal and regional variability of CO2 air‐sea fluxes

We examine the consequences of using four exchange coefficient versus wind speed parameterisations on the air‐sea CO2 flux estimation, using one year of satellite wind speed measurements at global scale. Estimates of the global net CO2 flux to the ocean vary from 1.2 to 2.7 GtC yr−1. Additionnaly, we show 1) that the relative importance of outgassing and absorption fluxes is dependent on the form of the relationships due to a correlation between the direction of the flux and wind speed and 2) that differences on global net air‐sea flux deduced from different K parameterisations are primarily due to parameterisation differences in the wind speed range between 4 and 17 ms−1.

The effect of local wind on seismic noise near 1 Hz at the MELT site and in iceland

Abstract The mantle electromagnetic and tomography (MELT) experiment on the east Pacific rise near 17°S was the first large teleseismic experiment on a midocean ridge. During the six-month deployment, no compressional arrivals were well recorded above 0.5 Hz. In comparison, the ICEMELT experiment in Iceland recorded compressional arrivals at 1-2 Hz from about 2 earthquakes per month. We compare noise spectra from the two experiments and show that this difference in detection is at least in part a result of noise. Near 1 Hz, seismic noise in the oceans is produced locally by wind-generated waves. At both experiment sites, 1-Hz noise levels are well correlated with local sea-surface-wind speeds derived from satellite observations. For a given wind speed, 1-Hz noise levels are about 10-20 dB lower in Iceland. At the MELT site, cross-correlations of wind speed with the logarithm of noise in a narrow-frequency band yield correlation coefficients exceeding 0.7 at frequencies between 0.4 Hz and 2 Hz. Noise levels at 1 Hz increase with wind by 1.3-1.4 dB per m/sec for wind speeds less than 10 m/sec. For the ICEMELT experiment, high correlation coefficients extend to markedly higher frequencies for coastal stations, and there is a 10-dB drop in 1-Hz noise levels 100-km inland. Noise levels increase by about 0.8 dB per m/sec. The strong correlation between wind speed and 1-Hz seismic noise provides justification for using satellite wind speed data to search for locations on the global spreading system where there is a better probability of recording high-frequency arrivals. The calmest sites are found on the northern east Pacific rise, near the equator in all oceans, and near 34° N and 22° S on the mid-Atlantic ridge.

Assessing the seasonality of the oceanic sink for CO2 in the northern hemisphere

Seasonal CO2 fluxes are estimated from quarterly maps of ΔpCO2 (difference between the oceanic and atmospheric partial pressure of CO2) and associated error maps. ΔpCO2 maps were interpolated from pCO2 measurements in the North Atlantic and the North Pacific Oceans using an objective mapping technique. Negative values correspond to an uptake of CO2 by the ocean. The CO2 flux for the North Atlantic Ocean, between 10°N and 80°N, ranges from −0.69 GtC/yr, for the first quarter (January‐March), to −0.19 GtC/yr for the third quarter (July‐September) using the gas exchange coefficient of Tans et al. [1990], satellite wind speeds, and a correction for the skin effect. On annual average, the North Atlantic ocean (north of 10°N) is a sink of CO2 ranging from −0.23 ± 0.08 GtC/yr (gas exchange coefficient of Liss and Merlivat [1986] with Esbensen and Kushnir [1981] wind field) to −0.48 ±0.17 GtC/yr (gas exchange coefficient of Tans et al. with satellite wind field). The CO2 flux for the North Pacific, between 15°N and 65°N, ranges from −0.66 GtC/yr from April to June to zero from July to September. For the Atlantic, the errors are generally small, that is, less than 0.19 GtC/yr, but for the Pacific considerably larger uncertainties are generated due to the less extensive data coverage. The northern hemisphere ocean (north of 10°N) is a net sink of CO2 to the atmosphere which is stronger in spring (April‐June), due to the biological activity, with an estimate of −1.23 ± 0.40 GtC/yr averaged over this period. The annual mean northern hemisphere ocean flux is −0.86 ± 0.61 GtC/yr.

Satellite sea surface temperature: a powerful tool for interpreting in situ pCO2 measurements in the equatorial Pacific Ocean

In order to determine the seasonal and interannual variability of the CO 2 released to the atmosphere from the equatorial Pacific, we have developed p CO 2 -temperature relationships based upon shipboard oceanic CO 2 partial pressure measurements, p CO 2 , and satellite sea surface temperature, SST, measurements. We interpret the spatial variability in p CO 2 with the help of the SST imagery. In the eastern equatorial Pacific, at 5°S, p CO 2 variations of up to 100 μatm are caused by undulations in the southern boundary of the equatorial upwelled waters. These undulations appear to be periodic with a phase and a wavelength comparable to tropical instability waves, TIW, observed at the northern boundary of the equatorial upwelling. Once the p CO 2 signature of the TIW is removed from the Alize II cruise measurements in January 1991, the equatorial p CO 2 data exhibit a diel cycle of about 10 matm with maximum values occurring at night. In the western equatorial Pacific, the variability in p CO 2 is primarily governed by the displacement of the boundary between warm pool waters, where air–sea CO 2 fluxes are weak, and equatorial upwelled waters which release high CO 2 fluxes to the atmosphere. We detect this boundary using satellite SST maps. East of the warm pool, Δ P is related to SST and SST anomalies. The 1985–97 CO 2 flux is computed in a 5° wide latitudinal band as a combination of Δ P and CO 2 exchange coefficient, K , deduced from satellite wind speeds, U . It exhibits up to a factor 2 seasonal variation caused by K -seasonal variation and a large interannual variability, a factor 5 variation between 1987 and 1988. The interannual variability is primarily driven by displacements of the warm pool that makes the surface area of the outgassing region variable. The contribution of Δ P to the flux variability is about half the contribution of K . The mean CO 2 flux computed using either the Liss and Merlivat (1986) or the Wanninkhof (1992) K – U parametrization amounts to 0.11 GtC yr −1 or to 0.18 GtC yr −1 , respectively. The error in the integrated flux, without taking into account the uncertainty on the K – U parametrization, is less than 31%. DOI: 10.1034/j.1600-0889.1999.00025.x

Satellite sea surface temperature: a powerful tool for interpreting in situ pCO2measurements in the equatorial Pacific Ocean

In order to determine the seasonal and interannual variability of the CO2 released to the atmosphere from the equatorial Pacific, we have developed pCO2-temperature relationships based upon shipboard oceanic CO2 partial pressure measurements, pCO2, and satellite sea surface temperature, SST, measurements. We interpret the spatial variability in pCO2 with the help of the SST imagery. In the eastern equatorial Pacific, at 5°S, pCO2 variations of up to 100 μatm are caused by undulations in the southern boundary of the equatorial upwelled waters. These undulations appear to be periodic with a phase and a wavelength comparable to tropical instability waves, TIW, observed at the northern boundary of the equatorial upwelling. Once the pCO2 signature of the TIW is removed from the Alize II cruise measurements in January 1991, the equatorial pCO2 data exhibit a diel cycle of about 10 matm with maximum values occurring at night. In the western equatorial Pacific, the variability in pCO2 is primarily governed by the displacement of the boundary between warm pool waters, where air’sea CO2 fluxes are weak, and equatorial upwelled waters which release high CO2 fluxes to the atmosphere. We detect this boundary using satellite SST maps. East of the warm pool, ΔP is related to SST and SST anomalies. The 1985–97 CO2 flux is computed in a 5° wide latitudinal band as a combination of ΔP and CO2 exchange coefficient, K, deduced from satellite wind speeds, U. It exhibits up to a factor 2 seasonal variation caused by K-seasonal variation and a large interannual variability, a factor 5 variation between 1987 and 1988. The interannual variability is primarily driven by displacements of the warm pool that makes the surface area of the outgassing region variable. The contribution of ΔP to the flux variability is about half the contribution of K. The mean CO2 flux computed using either the Liss and Merlivat (1986) or the Wanninkhof (1992) K’U parametrization amounts to 0.11 GtC yr-1 or to 0.18 GtC yr-1, respectively. The error in the integrated flux, without taking into account the uncertainty on the K’U parametrization, is less than 31%.

A comparison of ship- and scatterometer-derived wind speed data in open ocean and coastal areas

Quality controlled wind speed observations from merchant ships have been compared with ERS-1 scatterometer data. The ship and satellite wind speed pairs were well correlated at 120 km separation in the open ocean, reducing to 40 km in the North Sea. The maximum allowed separation of ship and scatterometer wind speed pairs had to be further reduced to 20 km in the North Sea to avoid matching of coastal ship wind speed data with scatterometer data from more exposed regions. Spurious biases in the comparisons were caused by the error variability of the scatterometer data (0.5 m s-1) being significantly less than that for the ships (2.0 m s-1). The unbiased regression was: U10n(ship)=1.025 U10n(satellite)+0.255 No significant enhancement of the scatterometer wind speeds occurred in the coastal region.