SPESP_Graphic Recommendations

	d	Graphic Recommendations	d	c
Thematic maps
		..
		This page contains: About cartography and infography Graphic recommendations (Prof. Calvo, Spain)
		.

		About cartography and infography Confusion in relation to both cartographic rationality and infographic creativity lies in the basis of mutual misunderstanding. This section stars from clarifying both concepts and then provides for a detailed reference of cartographic codes and rules that the Inforgraphic group produced (the Spanish Focal Point subgroup led by Prof. Calvo, Prof. Pueyo and Prof. Albisu) for the use on European Spatial Planning Cartography. Shortly speaking, one may say that cartography aims to provide objective representation of reality (maps) while infography aims to provide meaningful representation of human imagination (images). Infographic is, then, a more general concept since it may involve cartographic inputs and processes, if decided. Cartography, as a rational and objective representation of reality, follows a scientific-oriented methodology. As in any science, the method is more important than the final outputs since they are always provisory (a scientific answer can be rejected if wrong, but there is no way to assure permanent acceptance). Cartography is meaningful or misleading depending on the consistency of the data used as input and the process applied to transform it into a particular map. Anyway, maps always provide for a partial view (any scientific answer is never complete, but provoke new questions). Transparency in the data and formulation is required to make the overall method objective, so others can duplicate it getting exactly the same output. Infography may use cartographic products and methods as inputs, but they are free from following strict scientific rules. Creative methods follow an artistic-oriented procedure. While scientific methods always start from a pre-fixed question (and always provide partial answers involving further questions), artistic methods start from a solution to a question which is not clear or simply unknown. Creative methods invent symbols (so an alphabet) an rules to operate with the symbols (so a grammar). They do not discover the symbols “telling the true” from a systematic searching process: they invent them based on analogy or pure imagination. Creative methods are meaningful when they are efficient communicating, suggesting and evoking in the majority of readers the desired information. There is no need for transparency in how the symbols were invented. Creative methods use to be efficient communicating feelings and emotions (they can not avoid to communicate them somehow), but are risky communicating concepts or ideas when they are very specific. Creative methods may provide complete answers.
		Graphic recommendations (by Prof. Calvo, Spain) Resulting from the discussions of the Infographic’s group, a precise set of recommendations to other working groups in relation to how to map statistical data were produced (by Prof. Calvo, from the Spanish Focal Point). The maps elaborated to show the present state of the EU, to be used as a means of investigation or to display the results of research related to the different groups collaborating on the preparation of the ESDP, will always be a personal work that will have to be used in order to faithfully transmit that which the author wishes to underscore. In order to satisfy this premise the author of the cartography must establish visual codes according to the rules obeying graphic semiology in such a way that the reader can decode them avoiding errors of interpretation. Decision making, when it involves issues related to spatial planning, is never the result of considering a single variable, since spatial planning is holistic and integral itself. Therefore, the cartography developed for this purpose must maintain certain rules of internal coherency in managing the visual variables that, albeit very briefly, we attempt to point out on these pages. We understand that each map is a personal interpretation of a different reality and it has and should its author stamp and imprint.
		All that is intended by this document is to homogenize the language employed and to avoid the basic semiological errors arisen as a consequence of the different professional appointments. One of the main questions makes reference to the choice of colour ranges when is attempted to bring out values located above and below the average of the different reference units (points, lines or superficial layouts). Traditionally, graphic semiology assigns the range of warm colours to those values located above the average, since this range is understood as a salient respected to the plane of representation whereas the cool range is perceived as being more penetrating. This means that colours are not observed in a uniform way: greater importance is given to the warm range (yellows, reds, etc.) and they are situated more easily on the foreground than those of the cool range (blues, greens, etc.) Either option is possible, but we would recommend to use the warm range when representing those values -above or below the European average- which might be considered a problem and transmit the degrees of intensity by using different shades keeping the cool range for the reference units that are less of a problem. It is a question of playing with the “traffic-light-effect”, in which the yellow-reds indicate danger, whereas the blue-greens indicate that there is no obstacle.
		This idea is picked up in Figure 1 by establishing an average value, using the arithmetical mean, above and below this mean, 0.5 standard deviation intervals. Knowing that values between the mean and ? 1 standard deviation will account for somewhat more than 2/3 of the values (approximately 68%) and that close to 95% will fall within the mean ? 2 standard deviation range, that will leave approximately 5% above or below those values. This 5% will constitute the distribution extremes.


		The cartographic representation will therefore lead directly to the reference units with problems and will give less importance to those located within values that can be considered relatively normal. In interval distribution based on standard deviation is not always advisable, but it is a simple and easily homogenizeable way for the European whole. On other hand, by using the warm and cool colour range possibilities for representation and visual discrimination are broadened (five in the warm colour range and equal number in the cool colour range is suggested) which would end up being constrained in the event that only one of the two were used or with the use of shades of gray. In this case, it is recommended not to use more than five or six different shades, otherwise discrimination would be impossible. When a single range is used the most problematic spaces considered for spatial planning should have the greatest chromatic intensity, although in conventional cartography this greater chromatic intensity would always correspond to the more positive values. In case of a double colour range, we have already indicated the advantages of assigning the warm one to problematic spaces, representing the opposite with the cool range and maintaining the increment in values of shade intensity in all cases when separating the central values as shown on Figures 1 and 2, carried out as superficial and volumetric representations to demonstrate the visual results obtained. In order to be able to reconstruct the values referred dimensioning of visual variables should be obligatorily included in these maps. When choosing this type of representation one is consciously opting for that cartography which might be accused of being pessimistic, but it will undoubtedly be the best one for aiding decision making in highlighting problematic spaces. The indiscriminate use of different colours is one of the most usual mistakes which must be avoided. No semiological grading is possible between different colours. The grading is spontaneously established between “shades” of the same colour, although bi-chromatic and tri-chromatic grading could be possible maintaining specific fixed proportions of blue and yellow, and gradually adding red and any other colour, achieving in this way more harmonious values by having a harmonizing background of reference. Nevertheless, in the interest of simplicity and the need to unify criteria we suggest a series of primary colours, although small proportions (around 15%) of other colour, for example yellow, could be added to every interval, maintaining the shade grading, with the aim of achieving a greater harmonization of the entire series. The use of different figures such as, squares, triangles, circles, etc., is not advised, unless absolutely necessary to distinguish variables that are also absolutely different, especially when they can be represented by the visual variable of size, explained below. Using different figures to represent the different urban hierarchies of the cities is not advisable manner, but correct; we would suggest to use circles or spheres proportional to the population, introducing the hierarchy at a later point by using shades of colour. If treating with the selection of figures, circles or spheres are the most adequate as allow a better visual reconstruction of those figures that might be partially covered by overlapping. In any case, large sizes should always be placed in the background while smaller sizes be placed in the foreground. Size in the only visual variable that allows maintaining the relationship between the values of the real variable and those that are assigned in the visual variable. Representations of the size of the real variable can never be made through the use of colour or shade. If red or blue are chosen to represent N inhabitants of a NUT and split over the entire surface of the outlay unit will lead to visual messages that will induce error; these type of spot variables should not be represented onto a surface. Quite another matter would be in case of an outlay in raster cells of similar size, in which it would be possible to transmit the order of size through the shades of colour, but it would never be possible to transmit the real size variable. This visual variable can have a point, linear, superficial, or volumetric outlay, although what is finally perceived are surfaces. Larger or smaller points, flat figures which are more or less generous in their dimensioning, thicker or thinner lines, or figures representing volumes such as cubes or spheres that occupy space on a plane and therefore occupy surface, although they represent three dimensions. Figure 3 gives an idea of the differential possibilities of the representations by surface and volume.
		Therefore, bearing in mind that visual transmission is superficial, the proportionality between real and the visual variable must be sought in surfaces, except when dealing with volumetric variables, in which, theoretically, an idea of volume is being transmitted and we must therefore look for the correspondence of values of the real variable in the volumes it refers to, although their drawing will necessarily be in two dimensions. The presentation of the results must abide by the scale of presentation required which will be based on the number of individualized data. If the different figures are given a big dimension, they will overlap one another and one of the main values of the map will be lost namely that of simultaneously transmitting the general characteristics of the system as well as the particular data of each reference unit. If a reduced dimension is decided upon, there will be problems in being able to represent the smallest values unless a double constriction is used and the smallest values are assigned to the minimum perceptible size, maintaining proportionality in the rest of the distance covered by the variable. Consequently, the number of outlays to be made, the scale, and the distance covered by the real variable must be considered. When the variable covers a large distance, we suggest a volumetric representation that will allow the proportionality to be maintained and even endows a greater surface to the smallest values of the series, by which the representation can maintain an acceptable transmissivity even in such great distances as those that involve a cartography of all municipalities by population size. In order to do this, the rule is to introduce a square root in the relationship you wish to establish between the real variable and the longitudinal unit that will generate the volume of the representation, in the same way that it will be necessary to introduce a square root between both variables when you wish to establish a superficial representation. An appropriate way to present the key is the one indicated in the accompanying figure.
		The visual size variable is the only thing that can transmit the real importance of values such as population, the GDP, etc., which makes it a good reference as to the quantitative importance of the phenomenon on which other concepts can be introduced, resorting to filling the sphere or circle representing the size of the city or NUT with shade that speak of progressiveness or the scale order of different indicators. In this way, one can see the reference of the corresponding NUT, limited by the borders of its territory, the importance of its population, GDP or other variable indicative of size, and finally, the value grading of the indexes, that would be represented by the shade grading above or below the mean.

		Recommendations Projection GISCO, Lambert Azimutal Cartographic information: NUTSII, NUTSI, NUTSO Format DIN-A4 with graphic scale, without numeric scale indication, which will allow the information to be recovered, enlargement or reduction, adjusting as much as possible to the dimensions of DIN-A4 with a 1 cm edge. Map title and key should be put in the lower part of the paper. The borders of the different NUTS should be included, but with light lines indicating lesser hierarchical importance. If possible, do not use black for this purpose; gray is visually less weighty, it allows an easy recognition and it ties in well with any colour without subtracting force from the thematic part.