CASCADE was recently launched with a focused mission: to advance the understanding of critical urban health challenges faced by African cities and point towards practical and effective interventions.

Posted by & filed under CSAG student blog.

In (climate science) research, there are three basic components associated with the understanding a particular system of interest. First basic component is called “normal” pattern or trend; which is a series of events following a particular trend or pattern that is well known, understood and/or can be represented with high precision and accuracy. Second component is “noise” or outliers; which represents random and less important events existing somewhere in space of region of interest. Often, these events are ignored or neglected. Since they are somehow irrelevant and their absence do not hindered the understanding of system in question. Last component is anomalies; which represent events possessing some of the characteristics of both normal trend and outliers. Anomalies refer to series of events with their trend deviating from normal. Often, the presence of anomalies is due to addition of new element, modification of preexisting elements and/inability to capture all elements and processes in a system.

Scientific analysis could also be referred to as “pursued to pattern determination”. Over the years of scientific data analyses, it became tradition to scientists that there is a pattern almost within every data-sets, which defines the processes or elements interactions of a system. If not, the data represents no natural system, provided the approach of analyzing the data is valid. Thereof, an understanding of data-sets (the system) comes with determination/revelation of patterns. In many occasions, the first priority is to identify or find the “normal” pattern.  Once the pattern(s) have been defined (and declared as normal), eliminating the normal pattern from the data-sets reveal the existing anomalies and outliers. It is at that point that shift of interest occur – i.e. from focusing on (finding and understanding) “normal” patterns to finding and understanding anomalies. Commonly, the presence of anomalies highlights the gap(s) or limited understanding of the system. Climate system, for instance, comprises atmosphere, biosphere and hydrosphere. The limited knowledge with regard to each the above spheres presences anomalies (if not noises).

Ideally, if all the anomalies are revealed, understood and well presented, what will then be of interest on system (in climate science research)???

I came to believe that it the presence of anomalies keep keeps the flow of research or makes life interesting…

4 Responses to “Climate Science Research: Where is the interest in the absence of climatic anomalies…???”

  1. Greg Dor

    I personally dont think that we are in any danger of understanding all the possible anomalies of a chosen system, especially that of climate. The number of elements and interactions that go into determining how this system functions are to vast and complex to be able to predict or even understand completely. Thus there will almost certainly always be outliers to what is expected or understood to have happened. We may be able to gain further insight into why these anomalies have occurred but there will always be a different set of interactions that may throw up new anomalies. So for now, in our lifetime at least, I think we can be safe in the knowledge that there will always be something new to try and understand.

  2. Dean Harrison

    Without any disrespect intended, I feel as though this blog entry contains a number of conceptual errors.

    As discussed in Bruce’s first seminar, time-series are comprised of deterministic and stochastic elements.
    The stochastic element is the noise of the time-series with the random and unpredictable movement in the time series value. While it is not predictable, it is definitely an important element that is fundamental to decision making and cannot be “ignored” and is far from “irrelevant”. We need to have an appreciation of th degree of noise in order to understand the limit of certainty that can be placed on predictions.

    Outliers and extremes are far from irrelevant. Their importance depends on what question the researcher is attempting to answer. It may be the case that the researcher is only interested in the extremes and outliers. Think of disaster events. Eg flooding.

    The use of anomalies is just another way of framing a time series relative to a baseline or “normal” reference period. While these are deviations from the “normal” reference period, it does not mean that non-zero anomolies are noise. There will usually be some degree of a deterministic (predictable) element in the time series. When looking at climate variable time series, these anomalies could be due to internal variability that occurs at different time scales from seasonal variability to interdecadal variability. While they will comprise a stochastic element, there will also be a predictable pattern.

  3. Shakirudeen

    This is educational. As Gemma rightly said we cannot predict with certainty future events. However, based on certain parameters, we can give a range of predictions with accuracy. Nevertheless, i wish to add that projections are outputs of several parameters (fed into a system) rather than guess work. What makes it limiting are the anomalies and how they interact with the whole system.

    It is therefore, true that these deviations are what make work of scientists interesting.

  4. Gemma Bluff

    I agree that without anomalies and “noise” climate research would lack something, but I really don’t believe that we’ll ever fully understand the climate. Yes we may have an understanding of a large part of it, but this in itself is not 100% known. We may know how a system works, and the generalised expected happenings of it, but we’ll never know exactly how certain elements will interact at a particular time and in a particular place. For example, the Oklahoma tornado that touched down over the weekend apparently hasn’t occurred since 1999 (according to a number of news websites). How were the years in between then and now different?

    Furthermore, entering a 400+ ppm carbon dioxide state of the atmosphere, how are we to know what the climatological impacts will be – other than a guess work of predictions and projections (which have been stated as being “limited”)? We’ll never have a definitive answer to the climate, and will never know that on this day and this exact time “x” climatological phenomenon will happen. I think that’s what makes climate research such fun, and continuously interesting and educational, because you never really know whats around the next corner.