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. : 218-223, 2004 ISSN 0717-652X



Antonio Geraldo Ferreira & Paulo Travassos

Fundação Cearense de Meteorologia e Recursos Hídricos ­ FUNCEME Avenida Rui Barbosa, 1246, Bairro Aldeota, CEP 60115 ­ 221 - Fortaleza-CE Instituto de Ciências do Mar ­ LABOMAR Av. Abolição, 3207, Bairro Meireles, CEP 60165-081 ­ Fortaleza-CE
Universidade Federal Rural de Pernambuco - Depto de Pesca Av. Dom Manuel de Medeiros s/n - B. Dois Irmãos - CEP 52171-900 ­ Recife - PE


The objective of this paper is to present the "Atlas of the Sea Surface Temperature over the Atlantic Intertropical Basin by Using Data from the NOAA-14 Satellite (1995­2001)". In the Atlas, detailed information about the procedures utilized to retrieve SST from NOAA-14/AVHRR, over the Atlantic Ocean in the area encompassed by latitudes 10 N-20 S and longitudes 50 W-20 W, were compiled. Also included in the Atlas, is the complete set of monthly SST images and monthly historical average in the period between 1995 and 2001.Based on these images, short description of the SST patterns over the area mentioned before is made. The Atlas includes 2 CD's, one of which containing approximately 1250 images already processed, in jpg format, of the SST fields (daily, monthly and monthly mean), includies the software that enables to make the visualization, animation and some basic statistics over the images. The other CD contains the raw SST data base.



The Meteorology and Water Resources Ceara's Foundation - FUNCEME has a polar orbiting satellite reception system in order to receive images from NOAA/HRPT satellite series (National Oceanic and Atmospheric Administration/High Resolution Picture Transmission), that is located at the coordinates 3 44' S; 38o 34' W, in Forteleza city, Ceará State, Brazil. In 1993, FUNCEME initiated a research project in order to develop an operational model to retrieve Sea Surface Temperature (SST) using NOAA satellite images. At the end of 1994, after validation and tests phases, the model was installed and from 1995, it started to run operationally. Since then, until March 2001, the SST's, derived from NOAA-14 satellite data, over the intertropical Atlantic Ocean, was generated on a daily, weekly and monthly basis.

This information has been used in both climate and weather forecasts of the northeast region of Brazil (semi-arid region). In this region the quality of the main rainy season (period between February to may), is closely linked with SST patterns observed in the Pacific and intertropical Atlantic ocean. Therefore, SST fields derived from polar-orbiting satellites, in our case NOAA-14, can provide the regional coverage and spatial temporal resolutions, desired for the SST patterns monitoring and its impact over the climate of this semi-arid region.

The SST's generated have also assisted the REVIZEE (Living Resources in the Exclusive Economic Zone), a program implemented in Brazil, in 1994, to generate a database and knowledge about the exploration, exploitation, conservation and management of the living resources in the Brazilian Exclusive Economic Zone (ZEE). This program has as its main perspective to subside the sustainable use of sea resources over Atlantic ocean inside of the ZEE. The ZEE area, hereafter called ZEE-Northeastern, in which our analysis, concerning SST, is made, is delimited by the coordinates: latitudes 5° N ­ 15° S and longitudes 45° W ­ 20 °W.

This paper, therefore, describes the content of the "Atlas of the Sea Surface Temperature Over the Atlantic Intertropical Basin Using Data from the NOAA-14 Satellite (1995-2001)", that compiles detailed information of the procedures used to derive SST from NOAA-14/AVHRR, presents as well the complete set of monthly SST and monthly mean SST images, generated during the period between April 1995 and March 2001 over ZEE-Northeastern.


The images and products presented in the Atlas have been derived from the data supplied by the NOAA 14/AVHRR received at FUNCEME. Since that NOAA-HRPT reception system was acquired only with the satellite tracking and reception software, and not included any software for AVHRR data processing or products generation, it was necessary to develop all the basic processing procedures, in particular, for AVHRR data.

The set of images presented in the Atlas and in the CD were derived from the NOAA-14 polar-orbiting satellite, during the period from April 1995 to March 2001. The cover area defined for the products generated (channel 2 images, cloud masking images and SST's-daily, monthly and mean), was between the latitudes: 10° N-20° S and longitudes: 50° W-20° W. During the period from January to March-1995 and April to December-2001, the daily, weekly and monthly SST maps were not generated in function of operational problems in FUNCEME´s polar-orbiting reception system. The problems were solved only in January 2002 with the acquisition of a new reception system and data processing. For this reason, the Atlas and the CD contain only the SST maps generated in the period between April 1995 and March 2001.

Since 1995, after completing the validation and test stage, an algorithm for SST retrieval using NOAA-14 satellite data, was installed at FUNCEME. This algorithm involved the following procedures: Input:

NOAA/AVHRR-channels 1, 2, 4 and 5 Geolocalization Radiometric correction Cloud masking Atmospheric correction SST maps (daily, weekly and monthly).

The geometric correction (or navigation) is the process responsible for registering the image coordinates with the geographic coordinates (latitude, longitude), and it is made for each image (or channel) of the AVHRR. The geometric correction procedure extracts the orbital parameters of each tracking of the NOAA satellite series in order to establish the relationship between the pixels of the images and its respective geographic coordinates. The cloud masking detects pixels contaminated with clouds and separates them from those not contaminated in the NOAA/AVHRR images. This process must be an integral part of any algorithm that derives SST from satellite, since pixels contaminated by clouds cannot be used to retrieve SST. The cloud masking was implemented and calibrated, as presented by França and Cracknell (1995), for use in tropical regions. This method is based on 5 (five) different cloud masking techniques (Gross Threshold Technique, Q Technique, Coherence Technique-Partial Coherence Technique, New Coherence Technique and TB4-TB5 Technique), based on the spectral information of the AVHRR channels 1, 2, 4 and 5.

The correction of the atmospheric influences in the channels of the AVHRR sensor was made by using MCSST (Multi-Channel Sea Surface Temperature) algorithm, developed by NOAA, and the equation used to SST retrieval is expressed by:

TSM = ao Tb4 + a1(Tb4 ­ Tb5) + a2(Tb4-Tb5) (1.0/cos Ø -1) ­ a3

where: Tb4 and Tb5 are the brightness temperature in channels 4 and 5 respectively, q is the satellite zenithal angle for a given pixel in the channels 4 and 5 of the AVHRR, and the parameters ao, a1, a2 and a3 are: ao= 1.017342, a1= 2.139588, a2= 0.779706 and a3 = 278.43. These parameters are brought up to date periodically by NOAA.

The algorithms and the visualization software that compose the model for SST retrieval, was implemented in C++ language and uses a tool library called X classes, developed by FUNCEME's technicians. It allows excellent quality graphical presentation and has a user-friendly environment, as can be seen in Figure 1.

The process described previously can be better visualized through the sequence showed in Figure 2, where we can observe in Figure 2a, the channel 2 image from 03/20/2000 - 18:00 GMT after applying the navigation process. Figure 2b is derived from Figure 2a, after applying the cloud masking processes. After both process, cloud masking and atmospheric correction, Figure 2c shows the SST for 03/20/2000 - 18:00 GMT corrected. The monthly SST for March 2000 (Figure 2d), is generated making the composite of the daily SST images.

Figure 1. Main screen of the processing program of the SST/NOAA

Figure 2a. NOAA-14/A VHRR image - channel 2 - 03/20/2000, 18:00 GMT

Figure 2b. NOAA-14 image - channel 2 - 03/20/2000,
18:00 GMT on which the cloud masking was applied

Figure 2c. SST image derived from AVHRR for the day 03/20/2000 - 18:00 GMT

Figure 2d. SST map for March 2000, produced making the composite of the March daily SST maps

In order to generate the monthly SST maps, the algorithm developed searches, pixel by pixel, among the daily images that will go to compose the monthly SST, for that pixel which has the higher temperature value and stores this value. After a complete reading of the image and selection of the pixels with higher temperature value, the SST monthly map is generated. The performance of the algorithm that retrieves SST was made as described by Ferreira et al (1996).

Once the images are generated, they remain available to the FUNCEME´s forecasters who carry out the monitoring and forecast of the weather and climate for Ceará State and the northeast Brazilian Region (NEB), as well to the technicians that makes SST monitoring.

Concerning monthly mean SST´s fields, presented in the Atlas, it was calculated as follow: a computer program, written in C language, reads the pixels that have the same location in each SST monthly image, and then adds and stores these values. The following step consists of the division of these values by the amount of pixels involved with the same location in the images. This process is repeated for all pixels of the image. Thus, at the end of the processing, the NOAA 14/AVHRR monthly mean SST image is obtained.

If a given pixel is contaminated by cloud in all the images, it will appear as cloud in the resultant image. If a given pixel is contaminated by cloud in one image and not in another, the pixel contaminated by cloud is not considered in the average calculation. Finally, in order to make a qualitative comparison with the SST generated from NOAA 14/AVHRR, whose spatial resolution is 1 km x 1 km, the monthly climatology SST fields, both global and regional, over the Atlantic Ocean basin, plotted from the data supplied by the National Centers for Environmental Predictions-NCEP (spatial resolution: ~111 km x 111 km) and by Institut de Recherche Pour Le Développement-IRD (spatial resolution: ~222 km x 222 km), respectively, is presented in the Atlas.


The analysis of the monthly and monthly mean SST maps derived from NOAA-14/AVHRR, allowed better accuracy in the verification of the patterns of this physical parameter (SST) over the intertropical Atlantic Ocean, in the area covered by the images, if compared to those supplied by the global models, since that the spatial resolution of the AVHRR sensor in the sub-satellite point is 1 km x 1 km, while the global models are around 111 km x 111 km.

Analyzing the SST monthly images, in the period between April1995 and March 2001, we can verify that SST in adjacent areas of the Brazilian coast between the equator and the latitude 20o S varies between 26 C and 30 C, and in isolated areas can surpass this. However, in some months of the year, mainly July to September, between the longitudes of 1o S and 20o S, in the coastal zone and neighboring regions, SST reaches values between 25 oC and 26 oC. Inside of the ZEE-Northeastern the SST in the analyzed period vary during the year, between 23 oC and 30 oC, although in isolated areas were registered higher values than the last one.

In the analyzed SST maps, the months of March, April and May of 1996, and February to June of 1998, are remarkable. During these periods, between the latitudes 5o N-20o S and longitudes 50o W and 20o W the SST surpassed 28 oC in the whole area. In March and April of 1998, in some isolated areas within of the previously cited region, the SST reached values higher than 30 oC.

The monthly mean SST maps show that between the latitudes 10o N-20o S and longitudes 50o W-20o W presents, during the year, a relatively reduced thermal gradient of approximately 5 °C to 6 °C, with the SST varying between 24 °C and 29 °C. However, in this area, a well-defined seasonal variation can be observed, with water masses of the same temperature displacing in the southeast-northwest direction, in the littoral area.

In the period between December and May (the austral summer/autumn), the entire area inside of the ZEE-Northeastern area presents raised temperatures, that is, superior to 28 °C, mainly in the months February to April, when waters with temperatures between 29 °C and 30 C, surround all of the ZEE-Northeastern coast. This last characteristic was observed due to the spatial resolution of the AVHRR sensor. Starting in May, this hot water mass starts to be "pushed" northward, as a consequence of the water penetration with lower temperatures, approximately 25 °C, in the extreme Southeast (latitudes 18o S-20o S and longitudes 22o W-20o W). This gradual advance of cold water, from the south Atlantic, can be observed until September, when are temperatures between 22 °C and 23 °C is observed in this extremity.

At this time (August/September), having as a central reference the latitude 5 o S, the coast of the ZEE-Northeast can be divided, in two sectors: the north, surrounded by waters with temperatures between 27°C and 28°C, and the south, where the water temperature is only 1°C lower, (between 26°C and 27°C). This division can also be observed in a vast oceanic zone adjacent to these two sectors (Figure 3).

Starting in October (Figure 3), the colder waters with temperatures lower than 25 °C, start to return, giving space to the intrusion of water at warmer temperatures, above 28 °C, which will again in the apex of the austral summer, surround all of the Northern/Northeastern coast of Brazil, extending, also along the whole adjacent ZEE.

A spatial-temporal feature of the SST in the ZEE- Northeastern that deserves to be noted is that its thermal gradient is always perpendicular to the shoreline, with the SST diminishing from littoral zone toward the oceanic zone. This fact is directly associated with the direction of the shoreline, with the narrowing of the Atlantic basin in the tropical region, resulting from the advance of the Brazilian coast, and, evidently, to the dynamics of the ocean currents and to the displacement of the isotherms at different times of the year.

Therefore, we can verify that the SST data, derived from NOAA/AVHRR, can be used to analyze the characteristics of the SST annual cycle at a regional level over the intertropical Atlantic Ocean.

Figure 3. Mean SST map for September (daytime only) derived for NOAA-14/AVHRR (The average was made considering September 1996-2000)


The SST retrieval from NOAA 14/AVHRR, due to its cover area and spatial and temporal resolution, is a fundamental tool to monitor this important physical parameter in the Atlantic Ocean basin, because SST has a great influence, for example, in the rainy season quality of the northeastern Brazilian region (NEB).

This influence can be explained by the fact that SST patterns, in both the north and south Atlantic basin, modulate the positioning of the Intertropical Convergence Zone-ITCZ, the main atmospheric system causing rain in this region from February to May of each year. This fact shows the importance of research and development in understanding the ocean/atmosphere interaction to improve the quality of the weather and climate forecasts for the NEB. The SST assimilation data, retrieved from sensors on board satellites, in the weather and climate (global and regional) numerical prediction models as well. Moreover SST data can be used with data from ocean color instruments (such the CZCS-Coastal Zone Color Scanner or SeaWiFS-Sea-viewing Wide Field-of-view Sensor) to map regions of high productivity and to develop models for the dynamics of the phytoplankton populations. It has been shown that these SST images are also useful for defining the distribution of marine fish habitat conditions and improving the catch of certain species such as albacore tuna (Gurney et all, 1993).


The NOAA-14/AVHRR data described in this paper and copies of the "Atlas of the Sea Surface Temperature Over the Atlantic Intertropical Basin Using Data from the NOAA-14 Satellite (1995-2001)", can be obtained writing to the authors of this work.


The work related to the "Atlas of the Sea Surface Temperature Over the Atlantic Intertropical Basin Using Data From the Noaa-14 Satellite (1995­2001)" was supported by the REVIZEE/SCORE-NE, by the CNPq (National Council for the Scientific and Technological Development), by the General Coordination of the REVIZEE MMA/SMA/DEGAM and by the FUNCEME. The authors are very grateful for it.


Ferreira, A. G., França, G. B., Oliveira, R. S., 1996. Um modelo Operacional de Estimativa da Temperatura da Superfície do Mar (TSM) através do AVHRR/NOAA-14, IX Congresso Brasileiro de Meteorologia, 1, 575 - 578. [         [ Links ]1]

França, G. B., Cracknell, A. P., 1994. Retrieval of land and sea surface temperature using NOAA-11 AVHRR data in north-eastern Brazil, International Journal of Remote Sensing, 15, 1695-1712. [         [ Links ]2]

França, G. B., Cracknell, A.P., 1995. A simple cloud masking approach using NOAA AVHRR daytime data for tropical areas. International Journal of Remote Sensing, 16, 1697-1705. [         [ Links ]3]

Gurney, R.J, Foster, J. L., Parkinson, C. L., 1993. Atlas of Satellite observations related to global change, Cambridge University P ress, 470 pp. [         [ Links ]4]


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