The first data to consider relate to rainfall. Satellites measure rainfall by looking at the outgoing long wave radiation or OLR, which is measured in watts per square meter (W m-2). The amount of OLR emitted from clouds is determined by the cloud temperature; the lower the temperature, the lower the OLR. Because the satellite cannot penetrate clouds when measuring OLR it senses the temperature at the cloud tops.
The temperature of cloud tops depends strongly on cloud height; the higher the cloud, the colder its top. Low cloud temperatures of course mean condensation and possibly rain. Experience shows that clouds which exceed a certain height inevitably cause rain on the ground. This is the basis for rainfall measurement from satellites: Low OLR means low cloud top temperature, which means large cloud hight, which means large likelihood of rain on the ground.
The animation shows OLR anomalies and therefore has to be interpreted as departures from a reference average. This reference is itself a function of space and time, about which nothing can be learned from the distribution of anomalies. The only interpretation of OLR anomaly charts can be that positive OLR anomalies indicate less than average rainfall, negative anomalies suggest higher than average rainfall.
To begin with, watch the date on top of the figure and identify start and end of the time series. Once you can orientate yourself in space and time, verify that you can follow the events which are described below.
Light blue tones over most of the tropical Pacific Ocean at the beginning of 1997 indicate that the region received nearly average rainfall (OLR is within 30 W m-2 of its long term January average). The western tropical Pacific Ocean east of the Philippines and the South Pacific Convergence Zone received slightly above average rainfall (OLR anomaly near -40 W m-2). This situation continues until early March when rainfall is close to average nearly everywhere.
During mid-March the rainfall situation in the Philippines, Papua New Guinea and northern Australia changes from slightly above average to slightly below average. Although the departure from the mean is within 30 W m-2 either side of zero, the trend is significant. This becomes evident in early May when a centre of above average rainfall develops on the equator at the date line (180°).
The pattern of light blue in the west and yellow/orange at the date line dominates the next few months. The regions of anomalous rainfall (both higher than average and lower than average) grow during August, particularly along the Intertropical Convergence Zone. The most intense rainfall anomaly occurs in October, when high rainfall is observed east of the date line.
During the 1997-1998 El Niño event the anomalously high rainfall persisted for many months, extending eastward along the equator first and settling at 150°W in January 1998. The strength of the anomalies decreased slowly, but the rainfall centre near 150°W was still visible in May 1998.
During the remainder of 1998 rainfall over the tropical Pacific Ocean was close to average. Some anomalously high rainfall over Indonesia and north western Australia in October and November 1998 is probably related to cyclone activity.
The next animation shows sea surface temperature (SST) anomalies in °C. Satellites can measure SST over cloud free regions. The animation shows weekly SST anomaly patterns, which are produced by selecting cloud free regions from many images collected during each week.
At the beginning of 1997 SST is close to average at most locations. Higher than average SST develops along the South Pacific Convergence Zone during February, accompanied as we saw by higher than average rainfall (in accordance with the positive feedback between SST and rainfall).
A first indication for an El Niño can be seen in late April 1997 when SST anomalies in the extreme east on the equator exceed +4°C. This is a confirmation that the change in rainfall from slightly above average to slightly below average in the west during February and March was significant: It indicated the beginning of an ENSO event which sent a solitary equatorial wave towards America. In late April this wave begins to affect the upwelling by replacing the usually upwelled cold water with warm equatorial water.
The suppression of upwelling along the South American coast and on the equator intensifies during the following months. While the coastal upwelling along the coast of Chile recovers in September 1997, equatorial SST anomalies remain well above average between the date line and Ecuador, reaching values of +5°C and more in October/November 1997 when the highest rainfall anomalies are observed east of the date line. Equatorial SST anomalies remain high until about March 1998; they then retreat to the region off Ecuador where high SST is still observed in May 1998.
The remainder of 1998 sees a return to long term average conditions (small SST anomalies). The last two months are characterised by the development of lower than average SST along the equator (the result of strengthened equatorial upwelling), which indicates that at the end of 1998 the Pacific Ocean was not in an El Niño situation.
contact address: