Title: Observing and modelling the interaction between Indian Ocean, atmosphere and coastal seas (OMICS).

Background

We present here the background material for this project, using information on how one near-coastal regime of India behaves. This description, based on the third Satish Dhawan Lecture by Shetye (2004), will allow us to see what aspects of this behaviour we understand and what aspects we do not understand, leading us to a research agenda that makes coastal ocean forecasting its ultimate goal. The region we examine is off Goa on the Indian west coast with NIO located here, it has emerged as one of the best observed regions of the Indian coastal regime.

We begin with an overview of the temporal and spatial scales of motion that occurs in the ocean. At the low end of both the scales (see schematic) are wind waves, which have typical temporal and spatial scales of the order of seconds and metres respectively. Next come motions that are known as internal waves, whose lower temporal bound is determined by the local density stratification, i.e., by the Brunt-Vaisala frequency. Typically the lower bound on these motions is of the order of hours. The next set of important time scales are approximately half a day and a day, corresponding to semi-diurnal and diurnal forcing due to astronomical tides. These temporal scales are the most important in the shallow regime (depths ~ 10 m) along the Indian coast. The next important time scale is the inertial period, which is a function of latitude; off Goa, the inertial period is about two and a half days. The larger scales depicted in the schematic are introduced by the wind. They form a broad spectrum, with the temporal scales ranging from days to weeks. The largest of this band of scales is associated with the seasonal monsoon winds and has a range from from weeks to months. The largest of the scales depicted in the schematic refers to the global thermohaline circulation.

Illustration 1: Important scales of motion in the seas around India.

The tide gauge at Mumbai, one of the few gauges in the tropics that have a data record stretching back at least a century, provides an example of how these scales of motion look in a single data record. The tide gauge shows the sea-level variations due to tides, storm surges, the seasonal cycle associated with the monsoons, and slower variations that have a time scale of years to decades. Not all these time scales, however, matter for a forecasting system of the kind that interests us.

Illustration 2: The location of the observation platforms for the measurements during March-April 2003. (a) The north Indian Ocean. (b) The Indian west coast. The solid lines offshore mark the 50, 100, and 200 m isobath. (c) Large-scale map of the study region off Goa. The dotted lines mark the 10 and 20 m isobaths. The current-meter locations are shown by filled circles and are labelled by the first letter of the nearby town and the depth of the water column at the location. Thus, A10 (A20) represents the current meter in 10 m (20 m) water depth off Arambol, M20 the current meter off Mormugao, and C10 (C20) the 10 m (20 m) current meters off Colva. Sea level was measured at Verem using a Valeport tide gauge. The winds were measured using an automatic weather station located on the terrace of the National Institute of Oceanography in Dona Paula.

The scales that matter for a forecasting system are associated with three types of motion: surface wind waves (small temporal and spatial waves), tides and tidal currents (temporal scale of half a day and a day), and coastal currents with periods ranging from a few days to weeks. The following discussion is based on observations off Goa (see map). Wind waves were measured and their spectra computed once every three hours during June 1996 and May 1997 off Mormugao Bay; the depth of the water column at the measurement location was 23 metres. During this period, winds were measured (10-minute average of magnitude and direction) from the terrace of the NIO building. Coastal currents were recorded at five locations (see map) for a month during March-April 2003. During these two months, sea level was measured at five locations within the channel of the Mandovi Estuary and at six locations within the Zuari Estuary. The locations where the measurements were made stretched from the mouth of the two channels upstream up to where the tidal influence is felt. Simultaneously, as during June 1996 and May 1997,winds were measured from the terrace of the NIO building. We now describe each of these three types of motion as seen in these data.

Tides

One of the most visible manifestations of the dynamic nature of the ocean is the variation of sea level due to astronomical tides. Motion due to tides is strongly periodic and leads to upward and downward movement of the water surface. This up-and-down motion of the surface, or of sea level, is accompanied by horizontal currents that are also periodic. The tides along the coast of Goa are mixed, with both semi-diurnal and diurnal constituents contributing to the field.

Illustration 3: Sea level (cm) measured by the tide gauge at Mormugao port (top) and tidal currents (cm/s) off Mormugao (bottom). Note the rhythmic changes associated with the tides and how the tidal current vector rotates over a day.

Illustration 4: Data recorded by a current meter deployed off Mormugao, Goa during March-April 2003. The axes have been rotated before plotting and the velocity components (cm/s) plotted are cross-shore and along-shore. The cross-shore velocity component (observed and tidal) (top), the along-shore velocity component (observed and tidal) (middle), and the residual (de-tided) velocity components (bottom) are shown. The one-month record allows an accurate estimate of the tidal component of the current because the tide has precise periodicities; this allows us to remove the tide from the record and estimate the non-tidal, or residual, current. Note that the wind-forced current, given by the residual, also shows some periodicities, but these are not defined as precisely as for the tides.

Analysis of observations in the Mandovi shows that amplitudes of the major tidal constituents increase from mouth to head until an increase in the height of the channel bottom prevents the tide from progressing farther. The degree of amplification, however, differs from constituent to constituent. Each constituent of the tide in such a channel can be looked at as a wave that is forced at the mouth by the large-scale astronomical tide. The tide in the channel is a consequence of the propagating wave and a reflected wave generated within the channel. Both the incident and the reflected wave experience frictional decay and undergo changes because of the variation in the cross-sectional area of the channel from mouth to head. Hence, there are interesting dynamics in action, but there are no estuaries along the coast of India where this sort of dynamics has been studied. One reason for this lacuna is, of course, data records (on how the tide varies within a channel) that can permit a careful analysis (i.e., roughly a month) have not been available. Records shorter than a month are not suitable for such analysis and it takes considerable effort to generate such data at a reasonable number of locations in an estuary. This lacuna has prevented posing of questions related to the dynamics, with the result that the Mandovi and Zuari estuaries in Goa, owing to their proximity to NIO, are perhaps the best-studied estuaries of India.

Illustration 5: Sea-level variation in the Mandovi during March-April 2003.

Illustration 6: Salinity (psu) at a location in the Mandovi during spring (upper panel) and neap (lower panel) tide.

Sub-tidal motion

Since tidal motion has precisely defined frequencies, it is possible to remove accurately the tidal component from the observed current. This makes it possible to examine the variation that is not linked to tides. We now look at the kind of motion that occurs in near-coastal waters off Goa at frequencies lower than the tidal frequencies, i.e, motions having a periodicity greater than a day. This current, from which the tidal component has been removed, is called the de-tided or residual current. The de-tided along-shore current off Mormugao shows variations on time scales that vary from days to weeks. Analysis to determine correlation between de-tided currents, detided sea-level in the Mandovi and Zuari estuaries, and low-frequency winds (low-pass-filtered to remove diurnal variability, or the variability due to the sea breeze) measured during March-April 2003 showed that the sea level along the coast and in the Mandovi and Zuari estuaries rises when the along-shore current is northward and drops when the along-shore current is southward. This relation implies that the along-shore current is in geostrophic balance. The data show that the along-shore winds measured on the NIO terrace and the along-shore currents in the near-shore shelf regime off Goa are highly correlated at periods less than 10 days. Hence, at these short periods, the current off Goa (called West India Coastal Current or WICC) behaves like a typical wind-driven eastern boundary current. At periods greater than 10 days, however, the WICC was strongly influenced by winds farther south along the coast, a phenomenon called remote forcing (because the forcing winds are remote from the effect, the current, they produce). The remotely forced component was comparable to the locally forced component at periods greater than 10 days. Thus, as noted by earlier studies for the seasonal cycle (see the discussion on the state-of-the-art of numerical modelling in India), the WICC is not forced purely by local winds and is therefore not a typical eastern boundary current. At these high frequencies (periods of the order of weeks), we know very little about the WICC: it is only the seasonal cycle that has been mapped and only its dynamics have been investigated. For a forecasting system, however, we need to build our understanding of the variability of the coastal currents off India on time scales ranging from a few days to a few weeks, the along-shore component of this sub-tidal current being as important as the tidal component even in the near-shore regime of the continental shelf.

Surface waves

To anyone who is going out to sea for a few days either for pleasure or work, the most important question is : how rough is it going to be out there? The roughness depends on the kind of waves that are going to manifest on the ocean surface. These surface waves, called wind waves, are generated by winds blowing on the ocean surface. The most convenient form for representing them is the wind spectrum. Consider, for example, two typical spectra that have been observed off Goa, one on 8 July 1996 and the other on 4 April 1997 at 1500 hours. The winds were measured at the same time from the NIO terrace. These winds are typical of Goa. When the southwest or summer monsoon is not active, the wind field is dominated by the diurnal cycle due to the sea breeze, a phenomenon common along the entire coast of India. When the summer monsoon is active, the winds are much stronger and have no particular periodicity.

Illustration 7: Spectra of wind waves of Goa.

Illustration 8: Magnitude of winds (wind speed, m/s) measured at NIO, Goa during June 1996 to May 1997. The horizontal axis gives the date for each of the 12 months.

Towards a forecasting system for the Indian seas

Notice that the explanations given above are dotted with “perhaps” and “may be”. This uncertainty is because of a lacuna in the available literature. There is hardly any study that gives reliable observations, and links the observed waves to wind field over the north Indian basin. The literature does have reports on observations or results of model studies, but it is difficult to find a careful analysis that establishes a causal link between the observed wave field and the wind field that might have caused it. Such an analysis must have both observations and model studies, and they much aim at answering questions like: are there typical regions where the swells seen along the coast of India get produced? If yes, how far and large is the region? We will need to address these issues as we progress towards a coastal ocean prediction system.

We need data of this kind along other locations of the Indian coast. The observations would then provide the food for thought necessary to pose questions of the kind raised earlier. An observational database must emerge to describe what goes on in the Indian coastal region, and this must be followed by analysis to define why it goes on. Many such “whats” and “whys” would then pave a way to a reliable prediction system.

This lacuna is not limited to wind waves, but extends to the other motions discussed above.

Though we have a good data base on the tides and routine tidal predictions are made for several ports of India by the Survey of India, the predictions are limited to water level. INCOIS uses a tidal model to make tidal predictions in the Bombay High region, but it is evident that such predictions need to be continually tested with observations and extended to the entire Indian continental shelf.

A coastal prediction system also needs to take the estuaries into account because their banks are often the most preferred locations for human settlement. Many of our major cities are located on banks of estuaries. Hence, our estuarine channels offer issues relevant to the society, interesting dynamics, and considerable ease in setting up observing systems. In spite of these advantages, the estuaries remain little studied. It is perhaps time to ask the question: do we need a re-examination of our priorities?

The sub-tidal motion is important even in the near-shore regime of the continental shelf off India. The limited current data available suggest that the dynamics of these currents is not linked merely to local winds even at periods as short as a fortnight. We know little of the physical processes on the Indian continental shelf and slope at frequencies higher than the seasonal cycle. The present state-of-the-art in India, which we discuss separately in the context of the global experience, is not sufficient to put in place a reliable forecasting system for the Indian seas. We need observations that will help map the variability at these intra-seasonal frequencies and modelling studies that will build our understanding of these currents, just as was done for the seasonal cycle of the currents over the last two decades.

It is this science that has to underlie any forecasting system for the Indian seas that forms the subject of the X Plan programme of NIO. This particular project, one of several under the umbrella called Supra-Institutional Project or SIP, focusses on the physical processes, there being companion projects that look at other aspects.

Hence, as noted earlier, this component of NIO's SIP has two core objectives, both indispensable for building the science underlying a forecasting system. The first core objective deals with observations, providing the what, or the bedrock of this science. The second core objective deals with modelling, providing the why, or the mathematical and computational tools that a forecasting system will be built with. Together, these two objectives are expected to enable the assembling of a forecasting system for the Indian seas.

Back