For the design of a company irrigation system it is necessary to have an adequate basic cartography that highlights in detail the altimetry of the surfaces to be irrigated. In fact, having chosen the type of system (sprinkling, micro-irrigation) and the model of dispensing devices to be used, it is necessary to proceed with the tracing and subsequent hydraulic dimensioning of the network, which cannot ignore the knowledge of the morphology of the soil.
The plano-altimetric representation of the territory can be carried out through a detailed topographic survey, which traditionally can be conducted through the use of tools such as topographic GPS or distance meters. It is known, however, that the use of significant detail is particularly onerous in terms of time and cost. In such cases, very often reference is made to technical cartography in 1: 10000 scale which, however, contains few details for the purposes of the case, or to the more detailed cartography in 1: 2000 scale, whose availability is unfortunately still limited to only urban areas.
Accuracy level of topographic maps
Generally the cards take on specific names according to the scale adopted; usually they are distinguished:
- geographical maps for scales from 1: 1,000,000 down
- chorographic maps for scales from 1: 1,000,000 to 1: 100,000
- topographic maps for scales between 1: 100,000 and 1: 10,000
- separate technical cartographies in medium-scale (from 1: 10,000 to 1: 5,000), large-scale (from 1: 5,000 to 1: 1,000) or very large-scale (from 1: 1,000 to 1: 500) maps.
The accuracy and level of detail of the map are directly related to the desired scale. It is clear that a large scale needs higher accuracy than a very small scale. To indicate the accuracy of a map, we usually refer to two coefficients, called mp (mean planimetric error) and ma (mean altimetric error). These coefficients indicate the average errors in the position of a point on the paper obtained from a printed copy of the same printed on a non-deformable support.
In general, the mean planimetric error is established in a value between ± 0.2 and ± 0.5 mm, on the scale of the map; for example, in a 1: 10,000 scale map it would be ± 2 to ± 5m, while in a 1: 2000 scale map it would be ± 0.4 to ± 1m.
The average altimetric error, on the other hand, is fixed in a value between ± 0.02 and ± 0.2 mm on the scale of the map for the numerical heights written in full on the map detected in correspondence with details of the terrain, while for the heights obtained from the level curves, a value between ± 0.1 and ± 0.5 mm is fixed, always on the scale of the map; in fact at the scale of 1: 10,000 it would be from ± 1 to ± 5m, while at the scale of 1: 2000 it would be obtained from ± 0.2 to ± 1m.
Topographic survey with the Google Earth service
The plano-altimetric survey of a land where to install an irrigation system can also be achieved through the Google Earth service. Google Earth is software that generates virtual images of the Earth using satellite images, aerial photographs and topographic data stored in a GIS (Geographical Information System) platform.
Google Earth is also a three-dimensional graphic application that allows you to view aerial and satellite photographs of the Earth in a fairly high detail. In the main cities of the planet, the program is able to show images with a resolution of less than one square meter.
Google Earth can allow the localization of a point through geographic coordinates, addresses or simply by exploring the various regions of the planet by scrolling on the images of the globe with the mouse. Most of the satellite images of large cities are available in high resolution, so that buildings, streets and even cars can be clearly seen. The level of resolution depends on the importance of the place: if for large cities the resolution is particularly high, for most of the earth’s crust it is around 15 meters. Resolution refers to the maximum level of detail allowed from a planimetric point of view.
Google Earth has managed to obtain through the Shuttle Radar Topography Mission (SRTM) a digital model of elevation on an almost global scale from 56 ° S to 60 ° N. At present, the resolution of data globally is 3 seconds (about 90 meters) while for the United States alone it is 1 second corresponding to 30 meters. These data allow an accurate reconstruction of most of the world’s mountain ranges.
Google has also started a long time ago a campaign of cartography purchases (satellite images and GIS information) from major international suppliers including Tele Atlas, US Navy, Europe Technologies, DigitalGlobe and Cnes / Spot Image. The different providers have provided cartographic and photographic material for the different geographical areas that cover different periods of realization.
Since May 2006, Google has made data regarding Italy available, providing services such as the search for addresses, businesses and routes, which until recently were only enabled in England and the United States.
Google Earth accuracy: quality and reliability of the topographic survey
A research carried out in 2009 at the Department of Agroforestry Engineering and Technologies of the Faculty of Agriculture of the University of Palermo verified the usability of the Google Earth service for the plano-altimetric survey of an area of land.
During this research, the level of accuracy of the survey that can be obtained from the Google Earth service was assessed compared to traditional maps in 1: 10000, 1: 2000 scale or with those obtained through the detailed survey, which are usually used by technicians during the design of irrigation systems.
The study was conducted by comparing the values of the altitudes and slopes estimated starting from the data detected by the Google Earth software, with respect to those that can be determined with the available maps or those generated by the detailed survey.
The comparison concerned two areas falling within two farms in the Sicilian territory (Licata and Rapitalà), whose territories have a different morphology.
The Licata company is located in the municipality of Bompensiere (CL), in the Fontanazza district, north-west of the town and is identified on the 1: 25,000 scale IGM cartography at F ° 267 II NO Montedoro (Fig 5.1). The entire farm area is approximately 50 ha divided into different crop sectors. The morphology of the territory is characterized by the alternation of flat areas and areas with slopes ranging from 5% to 25%, with altitudes ranging from 170 m asl (near the Gallo D’Oro river) to 284 m asl (on the square in front of the company center).
The Rapitalà company is located in the municipality of Camporeale (PA) and is identified on the 1: 25,000 scale IGM cartography at Sheets 258 IV NW-NE-SW-SE Camporeale (Fig 5.2). The entire company surface is about 175 hectares of which 105 are planted with vineyards; the morphology of the territory is mostly hilly, with altitudes ranging between 300 and 600 m asl and average slopes of about 10%.
Methodology used
To detect the morphological characteristics of the two companies, first of all two detailed topographic surveys were carried out, on a scale of 1: 2000, with the aid of a NIKON DTM 310 wave distance meter, by means of which the plane coordinates were measured. elevation of the points of the farm surfaces (50 ha for the Licata company and 12 ha for the Rapitalà company). With regard to the Licata company, the support polygon was first identified and surveyed, consisting of a total of 10 vertices, of which 5 constituting a closed polygon and the other 5 resting on it, which was referred to a local reference system.
For the Rapitalà company, on the other hand, a closed polygon was detected consisting of 3 vertices, which were also referred to a local reference system. In both cases, the sides and angles of the polygon were detected several times in order to improve the accuracy of both linear and angular measurements. The polygon has therefore been compensated both from the angular point of view (in order to respect the geometric condition according to which the sum of the internal angles must be equal to the theoretical value [(n-2) × 200c], where n is the number of angles of the polygon), both from the linear point of view (so that the planimetric coordinates of the first point and the last vertex coincide). The planimetric coordinates were subsequently also referred to the UTM system of national cartography.
After having proceeded with the planimetric compensation of the polygon, the altimetric compensation was carried out, having previously verified that the altimetric closure error “T∆” was lower than the altimetric tolerance limit. Once the support polygon was detected and the plano-altimetric coordinates calculated, the company surface was then surveyed in detail.
Figures 1 and 2 respectively show the planimetric position of the vertices of the polygonal support for the Licata and Rapitalà companies, as well as the points detected within the company surface. Approximately 3 months of work were required to carry out the aforementioned surveys.
Figure 1 – Plan of the topographical points of the Licata company scale 1: 4,000
Figure 2 – Plan of the topographic survey points of the Rapitalà company, scale 1: 4,000
After having obtained the different plano-altimetric data concerning the two farms (detailed topographic survey in 1: 2000 scale, regional technical cartography in 1: 10000 scale and survey obtained from the Google Earth plug-in), we proceeded to the comparison between the values of the shares obtained through the various methodologies.
Considering that this comparison is made possible only by the existence of a single and shared reference system, it was necessary to convert the coordinates of the CTR (Regional Technical Cartography) (referring to the national Gauss Boaga system) and those of the detailed surveys (referring to the European system UTM-ED50) in the UTM-WGS84 coordinate system used by Google Earth. This coordinate conversion was carried out through the use of the TRASPUNTO computer software, whose algorithm is based on a system of linear equations:
NGauss-Boaga = a NUTM + b EUTM + c (eq. 1)
EGauss-Boaga = d NUTM + e EUTM + f (eq. 2)
where a, b, c, d, e, f are coefficients depending on the areas to which the points belong.
Having obtained all the coordinates referring to a single reference system, we proceeded, through the ArcView® software, to the superimposition of the 2D image obtained from Google Earth with the 1: 10,000 scale cartography and with the points of the topographic survey, in order to to verify the differences in the heights and slopes calculated with the different methods and to evaluate their extent. The overlapping of the images was followed in order by the following steps:
Phase 1: Digitization and transformation of the level curves of the cartography in scale 1: 10.000 in points and subsequent interpolation, in order to obtain a DEM file CTR (DEM of the regional Technical Cartography)
Phase 2: interpolation of the points of the detail survey to obtain the DEM file RT (DEM of the Detailed Survey)
Step 3: Creating the DEM file GE (DEM of the Google Earth service)
Phase 4: Creation of two point grids that cover the entire extension of the farms under study in order to compare the shares detected by the DEMs CTR with those obtained from DEM GE
Phase 5: Subdivision of company areas into smaller surfaces and tracing within them of sections on which to compare the DEM obtained (DEM CTR , DEM GE , DEM RT ).
The graphs shown below (Fig. 3 and Fig. 4) show, for the two areas under examination, the comparison between the heights of the land obtained with the Google Earth service and the regional technical cartography 1: 10,000.
Figure 3 – Comparison between the DEMCTR and DEMGE shares of the Licata company
Figure 4 – Comparison between the DEMCTR and DEMGE shares of the Rapitalà company
From a first comparison made between the quotas obtained from Google Earth and those obtained using the cartography (CTR) relating to the Licata company and the Rapitalà company, it is evident that for each point the values of the quotas obtained using the two procedures are very close between their.
The following figures (fig. 5 and fig. 6) show the error relating to the measurement of the heights for the various points examined:
Figure 5 – Error connected to the value of the shares estimated through Google Earth of the Licata company
Figure 6 – Error connected to the value of the shares estimated through Google Earth of the Rapitalà company
As can be seen from the observation of figure 5, the differences between the heights of the points obtained using the Google Earth plug-in and the corresponding ones obtained on the basis of the CTR, in 86% of cases are less than ± 5m, which represents the inherent error in determining the shares of the land on a map 1: 10,000. Furthermore, considering that the use of the DEM obtained from the GE plug-in involves an average absolute error in the determination of the shares equal to 2.5m, it is evident that in the case in question the use of this tool for the identification of a DEM leads to fairly reliable results.
As can also be seen from Figure 6, the differences between the heights of the points obtained using the Google Earth plug-in and the corresponding ones obtained on the basis of the CTR in 95% of cases were, also in this case, less than ± 5m . Furthermore, in the case examined, the absolute average error in determining the shares was equal to 2.3m, and therefore also in this case the results of the use of this instrument for altimetric analysis of the territory are quite reliable.
The results obtained therefore indicate the reliability of the DEM obtained through the Google Earth plug-in and the consequent usability of the heights of the points of the ground in studies that require medium-scale cartography (1: 10,000), such as those aimed at design of collective irrigation systems.
The figures below (fig. 7 and fig. 8) show, by way of example, the elevation profiles in some sections extracted from the two areas.
Figure 7 – Elevation of the terrain along section A
Figure 8 – Elevation of the terrain along section B
The examination of the figures shows that in general the elevation profiles of the terrain obtained using all three procedures are very close to each other, managing in general to interpret the actual morphology of the terrain. Only in some stretches, where the elevation profile is more rugged, are there greater differences between the altitudes obtained from the topographic survey and those detected by the Google Earth service.
From the comparisons made with the data of the detailed survey it emerged that the use of the DEM CTR in general it has a higher accuracy than that obtained with Google Earth. Furthermore, it was possible to ascertain, for the profiles examined, how the average of the differences between the heights detected by the detailed survey and those detected by the cartography is equal to 1.30 m, with a standard deviation of 1.03 m and a value maximum of 6.8 m; while the average of the differences between the altitudes detected by the detailed survey and those detected by Google Earth is equal to 1.82 m, with a standard deviation of 1.26 m and a maximum value of 7.46 m.
Conclusions
The results obtained showed that the average altimetric error on the altitudes obtained from the DEM GE it was approximately 2.5m, lower than that associated with the use of traditional maps which, with reference to a scale of 1: 10,000, is ± 5m. This circumstance leads to affirm the validity of the plug-in used, for the realization of the cartographic support, by the Google Earth software, through which it is possible, in real time and in the areas in which no traditional cartographic support is available, to obtain a reliable elevation of the terrain.
The work carried out has allowed us to see how, in comparison with the detailed survey, the technical cartography is almost always more accurate than Google Earth, even if it was observed that in 63.7% of cases the values of the quotas estimated through the Google Earth plug-ins are characterized by average errors of ± 3m, generally referring to the border areas of the companies examined or near watercourses where, notoriously, the different terrain models are less accurate.
Even the analysis conducted on the slopes of the land showed that, although differences in the values obtained using the different altimetric restitution methodologies were found, there is a fairly evident trend between the slope values obtained with Google Earth and those obtained through the use of the CTR or the detailed topographic survey, for all the areas examined.
