SciELO - Scientific Electronic Library Online

Home Pagelista alfabética de revistas  

Servicios Personalizados




Links relacionados


Gayana (Concepción)

versión impresa ISSN 0717-652Xversión On-line ISSN 0717-6538

Gayana (Concepc.) v.68 n.2 supl.TIProc Concepción  2004 


Gayana 68(2) supl. t.I. Proc. : 29-39, 2004 ISSN 0717-652X



M.A. Barbieri1,2, C. Silva1, P. Larouche3, K. Nieto1 & E. Yáñez1

1. Pontificia Universidad Católica de Valparaíso, Valparaíso, Chile. E-mail:
2. Instituto de Fomento Pesquero, Valparaíso, Chile, E-mail
3. Institute Maurice Lamontagne, Mont-Joli, Canadá. E-mail:


In order to verify the ability of RADARSAT-1 images to detect mesoscale oceanic features and the possible link of these oceanographic patterns with jack mackerel (Trachurus murphyi) distribution in the waters off central Chile, a project was developed as part of the Canada Centre for Remote Sensing Globesar-2 program. The combined use of simultaneously acquired RADARSAT-1, AVHRR sea surface temperature (SST), SeaWiFS Chlorophyll a concentration (Chl), TOPEX/ERS altimeter and ERS-2 scatterometer wind data greatly enhanced SAR imaging capabilities for the detection of oceanic features. Results show that detection of coastal wind-driven upwellings, eddies, frontal boundaries and phytoplankton blooms is possible using SAR imagery, given the proper environmental conditions. Results also suggest that jack mackerel distribution coastal resources is mainly associated with frontal boundaries, upwelling waters and high chlorophyll concentrations detected by the remotely sensing images.



In coastal environments, oceanic mesoscale features such as wind induced upwellings and phytoplankton production add to the complexity of the ecosystem leading to strong spatio-temporal variability, which may be the primary determinant of recruitment to oceanic and coastal fisheries. In the Chilean central-south coastal zone, stomachs analysis of jack mackerel (Trachurus murphyi) revealed that approximately 90% of their food corresponded to euphausids (zooplankton) that feed on phytoplankton (Córdova et al., 2000), the very first link of the marine food web. In a context of global climate change, it is expected that coastal ecosystems will experiences changes in their physical characteristics, which could in turn affect primary productivity and the whole marine food chain.

In order to monitor these changing ecosystem conditions, scientists must have access to tools allowing observations on a global scale. The use of remote sensing is now seen as a very good way to provide this information; should be confirmed through ground- truth validation. In this paper, we will address the remote sensing identification of coastal upwelling, eddies, frontal boundaries and phytoplankton blooms in waters off central Chile, which are a very important mesoscale oceanic features involved in primary production. At the present time these features could be observed through different sensors. Each feature on a given image may be the result of many causative mechanisms, only a few of them being related to the sought-after relationship between the observed feature and actual phenomena in the area at the time. The oceanographic interpretation of sea surface temperature (SST) maps provides a synoptic view of the SST field and furnishes relevant information about spatial and temporal evolution of upwellings, eddies, meanders and oceanic fronts. SeaWiFS chlorophyll a (Chl) maps are the primary tool by which phytoplankton can be observed. SST and Chl maps are useful tools to identify probable fishing grounds. Unfortunately, the use of the visible (SeaWiFS) and thermal infrared (AVHRR) spectrum means that cloud cover is a great impediment to the generation of time series.

RADARSAT-1 is the first operationally oriented radar satellite system capable of timely delivery of large amounts of data. Possible oceanic applications include: mapping mesoscale currents and regional circulation patterns; identifying frontal zones, internal waves, eddies, meanders, upwellings, shears, natural films and wind fronts. Unlike infrared radiometers and visible imaging systems, the high-resolution synthetic aperture radar (SAR) is unaffected by cloud cover and visible light conditions. At the same time a multisensor approach can facilitate in mesoscale ocean studies. Interpretation of oceanographic features can be strongly supported by the analysis of AVHRR SST and SeaWiFS Chl data and simultaneously acquired radar images (Johannessen et al., 1994; Shandley & Svejkovsky, 1999).

In this study, we analysed a set of RADARSAT-1 SAR, AVHRR SST and SeaWiFS Chl images of water off central Chile with the objective of investigating the utility of these images in the identification of mesoscale oceanographic features and the possible relation of these patterns with the distribution of pelagic resources.


Oceanographic characteristics of the study area

The study area is located in the central-south Ocean Pacific coast of Chile, between 33-36S and 71-7430'W (Figure 1). This region is affected by the Chile-Peru or Humboldt Current (HC) system. The HC has a mean width of 2,000 km and is slow (< 5 cm/s). The system is composed of a group of currents and countercurrents, superficial and deep which changes its position seasonally. In the surface layer are two currents distinguished flowing towards the north (coastal and oceanic HC branches) and two others flowing toward the south (countercurrent of Perú). The coastal branch of the HC, circulating in a northerly direction, has a continuous flow from the surface to an approximate depth of 300 m, with maximum speeds of 11 cm/s, biggest speeds are observed in autumn (Robles et al., 1980).

The area presents a high spatial heterogeneity and displays complex mesoscale structures. It is possible to distinguish in the area a series of semi-discrete water bodies with dimensions between 10 and 500 km and with identifiable properties. They are dynamically unstable and there is a density field associated to the currents. The main oceanic process produced in the study area are: wind-driven coastal upwellings (Fonseca & Farías, 1987), coastal thermal fronts (Barbieri et al., 1987), filaments and eddies (Cáceres, 1992). The recently upwelled waters suffer a quick transformation that as it moves forward toward the front without sinking, or a convective sinking takes place at the border of the front.

The Chilean upwellings in mid latitudes are characterised by a narrow fringe in comparison to similar events taking place in other mid latitudes. The upwelling processes are discreet events with duration from 4 to 10 or 12 days. The upwellings are located at coastal foci associated with points and capes, where intense winds occur. In the study area two upwelling areas have been identified in front of Topocalma and Nugurne tips (Fonseca & Farías, 1987; Vergara, 1992). In a recent study, Bello (2001) ratifies these two upwellings areas and indicates that toward the south the upwelling is more intense.

Coastal upwellings are a wind-driven physical mechanism by which cold nutrient rich subsurface waters are brought to the surface, helping start or sustain phytoplankton growth. They have a profound influence on the marine food chain. The cold nutrient rich coastal upwelling make this region very productive.

Figure 1: Study area.

Remote sensing data

In order to evaluate the capability of SAR to detect the effect of natural surfactants produced by phytoplankton and wind-driven upwellings, we used a set of RADARSAT-1 images, ScanSAR Narrow 1 (SCN1) beam mode were acquired during the coastal upwelling and bloom period. The RADARSAT-1-1 SCN1 images used in this investigation were acquired on March 23 1997 and March 20 1999 over the same coastal area in central Chile (Figure 1). The images were acquired during ascending passes and have a nominal area of 300 km x 300 km with a spatial resolution of 50 m. After the raw data were imported, an adaptive median filter of size 3*3 was applied to the images to remove as much speckle as possible while preserving high frequency features present. The next step was to make the geometric correction involving the collection of Ground Control Points (GCP) relating uncorrected data to geocoding data using the GCPWorks application. Finally, backscatter values profiles were extracted from images using line vectors located perpendicular to the coast.

NOAA AVHRR imagery for the study area were received and pre-processed at the HRPT station operated by the Remote Sensing Laboratory of the Catholic University of Valparaíso. All the image processing from raw Level 1B scenes to the final remapped SST product were done using a standard multi-channel algorithm (McClain et al., 1985). The AVHRR images, with a spatial resolution of 1.1 km, were obtained in the same day as RADARSAT-1 acquisition.

A SeaWiFS (Sea-viewing Wide Field-of-view Sensor) image covering the study was acquired on March 20 1999 and it was processed using the SeaDAS 3.3 software to produce Chl maps. To generate the Chl was necessary to: (i) select, order and download (by FTP online) the SeaWiFS LAC Level 1A image from Goddard DAAC database; (ii) process the data in the SeaDAS 3.3 in order to generate Chl Level 2 image products; (iii) perform geometric corrections and produce Chl map. The computation of the Chl image was based on the OC2 empirical algorithm (O'Reilly et al., 1998). The SeaWiFS image was obtained day of RADARSAT-1 images

Wind Scatterometer data were collected during the acquisition of SAR imagery. ADEOS NSCAT wind speed and direction were collected during March 23 1997. An ERS-2 AMI wind speed and wind barb image was collected from NOAA/NESCDIS (Chang, 1996) during March 20 1999. A Topex/ERS-2 altimeter image was used on March 20 1999 to represent the sea height anomaly in the study area. The data was collected from the Global Near Real-Time Altimeter on-line system of Colorado Center for Astrodynamics Research (CCAR), University of Colorado. The sea surface height anomaly and geostrophic velocity was calculated using a mean circulation derived from the Naval Research Laboratory (NRL) at the Stennis Space Center in Slidell Mississippi.

Fishery data

Data were obtained from the purse seine fleet fishing logbook records of fishing companies in the study area during the acquisition days of RADARSAT-1 images. The georeferenced pelagic fishery data corresponds mainly to jack mackerel (Trachurus murphy) catches expressed in catch per trip. The fishery data were superimposed on the remote sensing images in order to evaluate the possible associations. It should be said that during the days when the images were obtained, the ships operated only in the south area of the studied zone. So there is only information about south Constitucion (35°20´SS).


In the area of study in March of 1997 the SST image showed that the upwelling event was present in Point Nugurne (35° 57'S), Point Carranza (35° 36'S) and in Point Topocalma (34° 07'S). On March 19th , an upwelling event had started in the area going from south to north. On March 23rd at the coastal station in Panul (33° 33'S) there was an average wind of 2.8 m/s, and at the coastal Carranza an average wind of 4.1 m/s with turbulence indexes (TI) of 22 m3/s3 and 69 m3/s3,respectively, where TI is defined as the 3rd power of the wind speed, in order to evaluate the quantity of energy available to induce mixing processes (Bakun, 1987). Figure 2a shows the RADARSAT-1 SCN1 image acquired the 23rd of March 1997, under conditions of relatively high south and southwesterly winds (8.2-9.1 m/s). These winds are located off shore and acquired by the scatterometer. The image shows a reduction in backscatter, near the coast in the same area of upwellings observed in the AVHRR SST image (Figure 2b). It can also be noted that colder (12C-14C) subsurface water is upwelled south of Punta Toro moving away from the coast towards the west. The possible relation between SAR backscatter and AVHRR SST images was studied, a non-linear regression model was fitted (R2 is 0.5, p< 0.05) between the average backscatter and average SST data extracted from the images using a 10*10 km grid. A total of 224 values data were extracted after eliminating the wrong (AVHRR clouds) data. There is a relation between the low SST (smaller than 14C) and low backscatter values. This relationship could be result of the fact that the cooler coastal waters, the wind, shadow of the land, and the incidence angle. The SAR image also shows the presence of dark lines running along the coast; these features are frontal boundaries produced by the upwelling waters. An eddy produced by the persistent south and southwest winds is observed in the SAR image and also in the SST image; this feature is located between 3430'S to 35S with an approximate diameter of 55 km at to 50 km from the coast. The eddy contains warmer waters (15C-16C) than the surroundings (13C-14C). The spiraling shape of the eddy suggests an anticyclonic flow, therefore, a raising of water toward its center should be produced. The eddy displayed in the SAR can be interpreted as a manifestation of effect of the wind. This mesoscale feature is also observable in the AVHRR image. Other eddies are also weakly observed in front of Carranza (3550'S) in the mid-south part of the SAR image where the Chilean continental shelf is wider (Figure 2a). Spiraling shape of the eddies suggest a cyclonic circulation, indicating, a sinking of water toward their center. During this day, the pelagic (jack mackerel) fisheries catches were located mainly in the south part of the study area, in SST higher than 14C and associated to the external area of the frontal boundaries produced by the upwelling front (Figure 2a).

In the study area on the March 1999 images the upwelling centers were present in Point Nugurne, Point Carranza and in Point Topocalma. On March 16th an upwelling event had started in the area. On that day at the coastal station in Panul average winds of 7 m/s and at the coastal station Point in Carranza with average winds of 8.2 m/s. The 20th of March 1999, the SST image, showed that in the area the upwelling event was begining to decline (Figure 2d) (more image similar to 2 b and 2d were used), in the coastal station near San Antonio winds averages of 3 m/s and in point Carranza of 6.7 m/s with turbulence indexes of 21 m3/s3 and 296 m3/s3, respectively. Upwelled water plumes extended 60 km off Constitution and 40 km off point Topocalma. The winds information obtained from the scatterometer indicates that average winds on that day varied from 7.3 m/s in the south oceanic sector and up to 10.6 m/s in the north sector (Figure 2 c). The other RADARSAT-1 SCN1 image was acquired 20th of March 1999, during a period of high south and southwesterly winds (9.3-10.5 m/s) in the study area (Figure 2c). The image doesn't show too much mesoscale and small features due to the homogeneous sea surface caused by the strong winds dominant in the area, however reduced backscatter in some places near the coast are shown, these features are frontal boundaries produced by the upwelling waters. The presence of a thermal gradient with lower temperatures along the coast is visible in the SST images, associated with coastal upwelling driven by winds (Figure 2d). It its observed from the Chl images that in those areas in which upwelling events are present, the highest Chl levels were reported (over 4 mg/m3), reaching in coastal areas values greater than 10 mg/m3 until 45 mg/m3. Meanwhile, waters with Chl values around 1 mg/m3 reached 250 km in the area north of the zone of study (Figure 2e).The SeaWiFS Chl image shows an east-west spatial gradient in the study area with higher Chl found in the coastal regions associated with the upwelling waters (Figure 2e). An anticyclonic eddy is manifested in the SAR image presenting a high backscattering. This mesoscale feature is also visible in the AVHRR SST and in the TOPEX/ERS-2 sea height anomaly (cm) and geostrophic velocity (cm/s) (Figure 2f). The eddy is located between 3330'S to 3450'S with an approximately diameter of 165 km and contains warmer waters (16C-17C) than the surroundings (14C-15C).

Figure 2: Remote sensing images acquired during the project: a) RADARSAT-1 with NSCAT wind (m/s) and fishery data superimposed; b-d) AVHHR SST (C); c) RADARSAT-1 with ERS-2 wind (m/s), fishery data, profile lines; e) SeaWiFS Chl (mg/m3); f) TOPEX/ERS-2 sea surface height anomaly (cm) and geostrophic velocity (cm/s). Bathymetric contours for 250 m are also overlaid on the images.

Figure 3 presents the profiles A (105-km), B (130 km) and C (140 km) extracted from the SAR, SST and Chl images using line vectors located perpendicular to the coast. In Table 1 the variables descriptive statistics are presented for each profile. Results indicate that the winds speeds were statistically different in the three profiles, the highest being to the north of the zone of study (profile A with an average of 10 m/s), and diminishing towards the south, reaching 8.4 m/s on the average in profile C. Low backscatter levels are found near the coast. The highest backscattering is present in profile B with an average of 39.3 which is significantly different from the one present in profile A (average 34.4 m/s) and in the profile C (average 33.9 m/s). The SST pattern shows the cold waters are found near the coast and the warmers are offshore; in profile B the increase in SST from 12C-16.5C with a temperature gradient of 0.11 C/km is also produced until the 40 km offshore. An increase in the Chl (2-40 mg/m3) is observed in the area of cold upwelled waters (12C-14.5C) and lower backscatter values. The profile line B cross the anticyclonic eddy and the centre of the eddy corresponds to the maximum backscatter level, the warmers waters (17C) and the highest sea level of the area.

Profile C shows correspondence between the backscatter levels, SST and Chl data extracted. A drop in the backscatter level is produced 17 km offshore, in the upwelling waters and in an area of high Chl concentrations (16 mg/m3). Other two drops in backscatter, with correspondence of high Chl concentrations, are produced 60 km and 70 km offshore. These features detected in both SAR and SeaWiFS images can be associated with the presence of naturally occurring surfactants related with the biological productivity of upwellings waters. These ocean surface features are produced in the upwelled waters (11C-14C) visible in the SST image. The profile C crossed a frontal boundary generated 140 km offshore; a notable increase in the backscatter and SST (0.2C/km) levels is produced in this location.

The 14°C isotherm is considered an indicator of upwelling waters for that zone and time of the year (Bello, 2001). According to such criteria, upwelling waters reached 30 km off the coast in profile A, 20 km in profile B, and 63 km in profile C (Figure 3). In this coastal area, the high and medium values of Chl were present, reaching its maximum at 10 km in profile A, at 30 km in profile B and at 15 km in profile C. The relationship between Chl and the SST showed an inverse correlation coefficient of ­ 0.95 in profile A, -0.75 in profile B and ­ 0.74 in profile C.

In the coastal sector, where upwelling waters were present, the backscattering showed the lowest values. In profile A the backscattering reached a stable maximum at 50 km; this variable showed the same behavior of the winds, with an increment towards the ocean, reaching 10.4 m/s. In fact, both variables presented a positive correlation (0.68). In profiles B and C both variables showed a positive correlation of 0.56 and 0.78 respectively, but it is negative if the whole profile is considered in the areas of upwelling. In profile B the backscattering showed the highest values between 45 and 80 km off the coast; it is in this area where a gyre was observed in the Radarsat image (Figure 2c). In the coastal sector the highest values of Chl were observed, in an area in which the continental shelf is very narrow (a depth of 600 m at scarcely 20 km off the coast). At the same time, in the oceanic sector, the altimeter image showed the presence of an anticyclone gyre (Figure 2f). This could be contributing to the wide presence of moderate Chl levels of in the area north of the zone of study, extending 250 km to the west of the coast (Figure 2e). In profile C the maximum backscattering was reached at 45 km, diminishing thereafter while the wind builds up monotonically from 30 km off the coast.

During March 20 1999, the distribution of jack mackerel catches per unit effort is associated with the surrounding waters of the eddy in SST of 13C-14C and high (> 2 mg/m3) chlorophyll a concentrations related with the frontal boundary of the upwelled waters.

Figure 3: Profiles of radar backscatter, SST, Chl and depth extracted from images of March 20 1999 using line vectors A, B and C located perpendicular to the coast. Arrows indicate possible relations between backscatter, SST and Chl.


The areas of upwelling and eddies observed with the employed images and the wind data acquired through the scatterometer are fundamental in the interpretation of the SAR images. This information is relevant due the fact that the data in situ are scarce in the study area. The results show that in the upwelling areas there exists a low superficial temperature of the sea, a high presence of chlorophyll and low values of backscatter. Radar backscatter is modulated by winds, surface films and other mechanisms that directly affect the height and steepness of the Bragg ripples on the sea surface. In this work the results suggest the influence of the upwelling mechanism in the roughness of the sea, results which coincide with the ones pointed out by Nilsson et al (1995), Clemente-Colón and Yan (1999).

It is well known that changes in sea surface temperature can in fact produce significant changes in radar backscatter (Nilsson and Tildesley, 1995).The air-sea boundary layer stability is a function of the air-sea temperature difference. According to Clemente-Colón and Yan (1999) stability changes produce changes in the turbulent flow over the sea surface. Lower SST increases stability and decreases turbulence which, in turn, results in lower wind stress. Consequently, the generation of Bragg waves is reduced and a lower radar return should be observed over colder waters.

In this work it is thought that the index of turbulence (TI), -measured starting from the speed of the wind- is slower in the coastal areas where the upwelling is produced than in the oceanic sector where the winds are stronger and the backscatter increases. As a matter of fact, when comparing the images of March 23rd,1997 (Figure 2a) with those of March 20th, 1999 (Figure 2c) it is observed in the upwelling areas that the backscatter was less in the first one when the winds and the index were lesser. When analyzing the last image a spatial variation of the speed of the wind is observed. (Figure 2c). Then, when analyzing the correlation (r) of the latter with the backscattering in the areas of upwelling in the profiles shown in the Figure 4, it is observed that such correlation is significant (from r = 0.56 to r =0.78) w for profile B and C.

Thus in profile A (Figure 3) when the wind increases the backscattering also increases along the whole profile. It is worth mentioning that the speed of the wind was stronger than10 m/s and a homogenization is produced in the Radarsat image (Figure 2c).

Theory indicates that radar backscatter has also been shown to depend on water temperature through changes in water viscosity, lower temperature increases water viscosity, increased viscosity produces increased damping and affects the initiation of Bragg waves, which also results in lower radar (Zheng et al, 1997). Besides, Clemente-Colón and Yan (1999) point out that the effect of changes in surface tension and dielectric constant with temperature are less significant than those caused by stability or viscosity changes.

The wind speed is a particularly important factor for SAR radar image of the water surface. As shown in other experiments, the sea surface is homogenized when the winds are too high. In our study, winds were above the generally accepted threshold of 8 m/s for slick detection and under the threshold of 12 m/s for detection of eddies and frontal boundaries (Johannessen et al., 1994).

Theory predicts that observed areas of low backscatter were caused by biological surfactants released by the shoreline productivity, by the phytoplankton blooms and accumulated in the convergence zone (upwelling waters) of the local current field (Beisl et al., 2000, Shandley & Svejkovsky, 1998; Beal et al., 1997; Trivero et al., 1998; Gower, 1994). Our observations thus tend to support the hypothesis of Johannessen et al. (1994) that states that phytoplankton generates natural oils which, given the proper environmental conditions, could affect the sea surface roughness and thus be detected by radar remote sensing as dark surface slicks. Our results indicate that zones of high chlorophyll a concentration or phytoplankton blooms are located in the upwelling area and are associated with low backscatter values. Some drops in the backscatter level correspond with increases in the Chl concentrations, possibly produced by dark surface of natural slicks.

The coastal upwelling produces slicks and areas of reduced radar backscatter, biological sources and shoreline wind sheltering compete in the sense that they, too, produce slicks and lowered levels of radar backscatter near the shore ­ even in the absence of coastal upwelling. In this work, in both cases, the zone was in presence of upwelling (Figure 2b and 2d) so we believe that the mechanism dominating the image is the first.

The observed backscatter variations may result from a variety of processes such as wind stress discontinuities, wave-current and accumulation of surfactants through surface convergence (Johannessen et al. 1994). Numerous examples have shown that warm currents have a stronger radar return than the surrounding colder water leading to a possible identification of their boundary (Hayes, 1981; Topliss et al., 1994; Askari et al., 1993; Nilson et al., 1993; Beal et al., 1997). Figures 2 and 3 show that the sea surface temperature difference between the coastal upwelling and offshore areas was about 4-6C, so we can suggest that the thermal discontinuities or the naturally-occurring surfactants were responsible for the large number of alongshore bands in the SAR images.

Coastal upwellings are a physical wind-driven mechanism by which nutrient rich subsurface waters are brought to the surface, helping start or sustain phytoplankton growth and have a profound influence on the marine food chain. These upwellings are normally detected using thermal infrared images but, once again, cloud cover often affects the use of this methodology. Theories also suggest that coastal upwelling can be detected using synthetic aperture radar (SAR) data based on their effect on the boundary-layer air-sea interactions (Weissman et al., 1980; Johannessen et al., 1996; Beal et al., 1997). Our results indicate that the upwelled cold waters produce lower backscatter levels than warmer waters. Applications of this methodology would supplement complete time series of coastal upwelling for regional ecosystem studies.

In the central zone of the area an anticyiclonic eddy is observed in the image of the altimeter. These eddies may be produced by the stress of the surrounding wind and a sinking of the thermocline. In a slightly coastal position an eddy is also observed in the RADARSAT-1 Scan Narrow-1 image that has a resolution of 50 m. The scarce resolution of this type of image limits its study and it hampers the detection of mesoscale oceanic features; besides, the influence of the Humbolt current of the coastal branch has a slow flow. The superficial features observed in the images may be either answers to oceanic phenomena or be reflections of meteorological phenomena.

An acoustic survey was carried out during May of 1999 for the estimation of abundance and biomass of small pelagic fish stock in the Chilean central-south zone (Córdova et al., 2000). Relationship was found between the jack mackerel distribution and the horizontal salinity gradient, the euphausids (zooplankton) concentration and the integrated Chlorophyll a concentration (from surface to 100 m depth). Jack mackerel was found in waters with integrated chlorophyll from 13.2 to 100.8 mg/m3, and consumes euphausids and lantern fish preferably. On the other hand, in the area where the anticyclonic eddy was observed, a high salinity and temperature and low oxygen values were found, but jack mackerel presence was not detected. It was found in the coastal external boundary of the eddy, next to the external boundary of upwelling waters, where the eddy presents a height about -2 to +2 cm.

Regarding the ocean features here studied, cloudiness is the main obstacle in using images to study temperature and color, as wind is ford RADARSAT images. The jack mackerel is a resource that presents a wide spatial distribution offshore. In that area winds may reach 10 m/s, as a consequence, its use may be limited. Nevertheless, for more coastal resources, such as sardines and anchovies or in areas where the intensity of the wind is slower, they should be explored and carefully studied in conjunction with in situ surveys, giving accurate spatial and temporal information on the distributions of important physical, chemical and biological variables and allowing determination of the important causative mechanisms modulating the SAR, AVHRR and optical and microwave imagery.


RADARSAT-1 images of the Chilean central-south zone outline mesoscale oceanic features such as wind-driven, waves and upwelling, phytoplankton blooms, frontal boundaries and eddies identified also on the AVHRR SST and SeaWiFS Chl images. This results demonstrate that RADARSAT-1, AVHRR and SeaWiFS manifestations of, respectively, surface roughness, sea surface temperature and chlorophyll a concentration can be compared in the study area, since the cold upwelled waters and high Chl concentrations are expressed as low backscatter levels in the RADARSAT-1 image. Nevertheless, the high intensity of the wind is a limitation in the use of these images for the study of the oceanographic features. High winds also hamper the establishment of an operative system of applications of images requiring the use of daily or weekly multitemporal images.

An anticyclonic eddy with an approximate diameter of 165 km is observed in the March 20 1999 RADARSAT-1 image and also in the AVHRR SST and TOPEX/ERS altimeter data. Survey results have demonstrated that the distribution is of jack mackerel is associated with the occurrence of frontal boundaries, upwelling waters, and high chlorophyll concentrations. In this study we have shown that these phenomena may be observed in remotely sensed images.

Satellite images obtained by radar sensors complemented with Chl and SST images, are useful tools that help to describe the jack mackerel distribution when it is located in the coastal surroundings and it could be employed in the operation of fishing fleets, whereas in the ocean where winds go higher than 10m/s, the use of the RADARSAT images present some limitations.


The authors would like to acknowledge the important contribution of the Canada Centre for Remote Sensing Globesar North-South University Linkage program to the realisation of this project. Special thanks to Mike Manore, Connie Johnson and Chris Hutton for their help with the project organization.

We would also like to thank the anonymous editors because of the valuable help and dedication to revise this manuscript. Our special thanks to Mr. Sergio Salinas of Pontificia Universidad Catolica of Valparaíso for his suggestions and to Mr.Martín Farías of the Pontificia Universidad Católica de Santiago for his encouragement in the development of this work.

A CONICYT-FONDEF D98-I022 project provided part of the funds necessary to carry out this research.


Askari, F., G.L. Geernaert, W.C. Keller & S. Raman. 1993. "Radar imaging of thermal fronts", Int. J. Remote Sensing, Vol. 14, pp. 275-294. [         [ Links ]1]

Bakun, 1987. "Monthly variability in the ocean habitat off Peru as deduced from maritime observations, 1953 to 1984. In: Peruvian Anchoveta and Its Upwelling Ecosystem: Three Decades of Change, D. Pauly & Y. Tsukayama (Eds.), IMARPE, GTZ-ICLARM, 46-74. [         [ Links ]2]

Barbieri M.A., F. Naranjo, E. Yáñez, M. Farías, G. Danneri & P. Rojas. 1987. "La pesquería artesanal del atún aleta larga en la zona de Valparíso y el satélite NOAA". Invest. Mar., Valparaíso, 15: 41-61. [         [ Links ]3]

Beal, R.C., V.N. Kudryavtsev, D.R. Thompson, S.A. Grodsky, D.G. Tilley, V.A. Dulov & H.C. Graber. 1997. "The influence of the marine atmospheric boundary layer on ERS 1 synthetic aperture radar imagery of the Gulf Stream", Journal of Geophysical Research, Vol. 102, pp. 5799-5814. [         [ Links ]4]

Beisl, C.H., F.P. Miranda & C.L. Silva. 2000. "Combined use of Radarsat-1 and AVHRR data for the identification of mesoscale oceanic features in the Campos Basin, Brazil", Proceeding of the sixth International Conference on Remote Sensing for Marine and Coastal Environments, Charleston, South Carolina, USA, 1-3 May 2000, Vol. 2, pp. 415-420. [         [ Links ]5]

Bello M. 2001. "Análisis de la Variabilidad Espacial y Temporal de los Eventos de Surgencia Costera en la Zona Central de Chile (32 - 36 S, 71 - 75 W)", Tesis para optar al título de Oceanógrafo, Escuela de Ciencias del Mar, Universidad Católica de Valparaíso, 64 p. [         [ Links ]6]

Bernal, P. 1990. "La oceanografía del sistema de corrientes de Chile-Perú en relación a las pesquerías pelágicas: Una revisión". In: Perspectivas de la Actividad Pesquera en Chile, M.A. Barbieri (Ed.), Escuela de Ciencias del Mar, UCV, Valparaíso: 35-48. [         [ Links ]7]

Cáceres, M. 1992. "Vórtices y filamentos observados en imágenes de satélite frente al área de surgencia de Talcahuano, Chile central". Invest. Pesq. (Chile) 37: 55-66. [         [ Links ]8]

Chang, 1996. NOAA/NESDIS, Office of Research and Applications, Oceanic Research and Applications, World Wide Web site ( [         [ Links ]9]

Clemente-Colón, P. & X-H. Yan. 1999. "Observations of east coast upwelling conditions in the syntheetic aperture radar imagery" IEEE Trans. Geosci. Remote Sensing. Vol. 37, pp 2239-2248. [         [ Links ]10]

Córdova, J., M.A. Barbieri, S. Nuñez & V. Catasti. 2000. "Evaluación acústica del recurso jurel en la ZEE de Chile", Fondo Investigación Pesquera, FIP N99-03, 200 p. [         [ Links ]11]

Fonseca T. & M. Farías. 1987. "Estudio del proceso de surgencia en la costa chilena utilizando percepción remota". Investigación Pesquera, Chile, Vol. 34, pp. 33-46. [         [ Links ]12]

Gower, J.F. 1994. "Mapping coastal currents with SAR using naturally-occurring surface slick patterns", Proceedings of the second ERS-1 Symposium, Hamburg, Germany, 11-14 October 1993, ESA SP-361, pp. 415-419. [         [ Links ]13]

Hayes, R.M. 1981. "Detection of the Gulf Stream", Spaceborne synthetic aperture radar for oceanography, The John Hopkins Oceanographic Study number 7, Beal, R.C., P.S. DeLeonibus and I. Katz (Eds.), pp. 146-160. [         [ Links ]14]

Johannessen, J.A., R.A. Shuchman, C. Wackerman, G. Digranes, D. R. Lyzenga & O.M. Johanessen. 1994. "Detection of Surface Current Features with ERS-1 SAR", Proceedings of the second ERS-1 Symposium, Hamburg, Germany, 11-14 October 1993, ESA SP-361, pp. 565-569. [         [ Links ]15]

Johannessen, J.A., R.A. Shuchman, G. Digranes, D. R. Lyzenga, C. Wackerman, O.M. Johanessen & P. Vachon. 1996. "Coastal Ocean fronts and eddies imaged with ERS-1 synthetic aperture radar", Journal of Geophysical Research, Vol. 101, pp. 6651-6667. [         [ Links ]16]

Nilsson, C.S., P. Tildesley & J. Peterson. 1993. "Mapping the East Australian Current with ERS-1 SAR and Shipborne Studies", Proceedings of the first ERS-1 symposium, Cannes, France, 4-6 November 1992, ESA Special Publication 359, pp. 95-100. [         [ Links ]17]

Robles F., Alarcón E. & Ulloa A. 1980. "Water masses in the northern Chilean zone and their variations in the cold period (1967) and warm periods (1969, 1971-73)", Proceedings Of the Workshop on the Phenomenon known as El Niño, UNESCO, Paris: 83-174. [         [ Links ]18]

Shandley, J. & J. Svejkovsky. 1999. "Observations of upwellings-related plankton blooms with AVHRR and SAR imagery", Proceeding of the fifth International Conference on Remote Sensing for Marine and Coastal Environments, San Diego, California, USA, 5-7 October 1998, Vol. 2, pp. 201-213. [         [ Links ]19]

Topliss, B.J., T.H. Guymer and A. Viola. 1994. "Radar and Infrared Measurements of a cold Eddy in the Tyrrhenian Sea", Int. J. Remote Sensing, Vol. 15, pp. 1173-1188. [         [ Links ]20]

Trivero, P., B. Fiscella F. Gómez & P. Pavese. 1998. "SAR detection and characterization of sea surface slicks", Int. J. Remote Sensing, Vol. 19, N3, pp 543-548. [         [ Links ]21]

Vergara, J. 1992." Simulación de un evento de surgencia costera a los 29,5S, Chile". Invest. Pesq. (Chile) 37: 89-96. [         [ Links ]22]

Weissman, D.E., T.W. Thompson & R. Legeckis. 1980. "Modulation of sea surface radar cross section by surface stress: Wind speed and temperature effects across the Gulf Stream", Journal of Geophysical Research, Vol. 85, pp. 5032-5042. [         [ Links ]23]

Zheng, Q., X-H. Yan, C.-R. Ho, V. Klemas, Z. Wang & N.-J. Kuo. 1998."Costal lee waves on ERS-1 SAR images" J. Geophys. Res.vol 103,n° C4, pp 7979-7993. [         [ Links ]24]

+ Globesar-2 Chile#9 project, Canada Centre for Remote Sensing.

* Fondef Efisat D98-I-1022 project, CONICYT.


Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons