| GPS Q&A
By Catherine Mansfield Q.I am interested in using a GPS/GIS data collection system to collect data for my GIS. How can I ensure the data collected is accurate? And are there tools that help me to understand the accuracy of the data collected? A.Accurate data is of a concern to most GIS users. Specifying the accuracy requirements and then understanding whether or not the data collected meet those requirements is important.
When discussing the accuracy of data in GIS there are really two areas to consider:
¥ the accuracy of coordinate data, and
¥ the accuracy of the attribute values.
Often when considering GPS / GIS data collection systems as a tool for data collection it is easy to focus more on the position accuracy and less on the attribute accuracy. As a GIS is a compilation of tools that allow you to not only create maps but also more importantly perform research and analysis and use as an aid in decision making, it is imperative that both the coordinate and the attribute data meet specified accuracy requirements.
To ensure that the data collected with a GPS / GIS data collection system meet the accuracy requirements for your purposes, it is necessary for you to first specify what the accuracy requirements are. It is then necessary to establish how you can ensure that the accuracy requirement is met. The following discussion considers both issues. The Position Data
Specifying the accuracy requirement for positions in the GIS will help determine what system that you use for the data collection. Today there are many GPS / GIS data collection systems on the market. The specifications on these systems vary from 5-meters to 1-centimeter. As a general rule, when the accuracy capability of a system gets better, the cost goes up. However, so to does the need for more rigorous field data collection techniques and data processing in the office to ensure that the specified levels are being met. Basically, the more accurate you want the coordinates to be, the more it will cost in terms of time, money and understanding.
Requirements for the positions may be specified in terms of required levels of confidence. Examples of this are:
¥ 68 percent of the data are sub-meter
¥ 95 percent of the data are better than 5m
¥ 99 percent of the data are sub-foot
The actual level which you select will depend on the application and use of the data in the GIS both today and the potential uses you can see for it in the future. Legal issues surrounding the use of the data may also define the required level of accuracy.
When using GPS / GIS collection to gain positions for your GIS there are a number of items that must be considered so that you can ensure you are getting the best possible performance from the GPS receiver. It is important to understand that GPS settings such as dilution of precision, signal-to-noise ratio, and elevation masks, the number of satellites to be used and the type of GPS data (code or carrier phase data) collected will affect the results you get. You should use the recommended settings in the documentation for the system.
Data collection field procedures are also important, especially if you desire higher levels of accuracy. As stated before, the more accurate the position data are required to be, the more rigorous the field data collection will need to be. One very important consideration is how far you can work from a base station and still achieve the system's specified accuracy. As a general rule it is possible to work at greater distances from a base station if you are collecting code phase data - these distances may be as great as a 1000 kilometers. With code phase data you can usually achieve accuracy specifications in the range of sub-meter to 5-meters depending on the GPS receivers specifications. If you are collecting carrier phase data you can achieve better levels of accuracy but you will need to work much closer to the base station. Distances for carrier phase data collection may only be as great as 10 kilometers for centimeter level accuracy and can extend out as far as 50 kilometers for accuracy better than 50 centimeters.
Most professional GPS / GIS data collection systems will specify a PPM value (Parts Per Million, i.e. 1 millimeter per 1 kilometer) as part of the accuracy specification. The PPM value refers to the degradation with distance from the base station, so that you can understand the level of accuracy you can potentially achieve as you work further away from a base station.
Distance from the base is not the only field procedure you need to be aware of. As the level of required accuracy goes up, so to will the need to mount the GPS antenna on a tripod, bipod or range pole. It is probably not valid to claim 30-centimeters or better with a unit and to be carrying the GPS antenna on your back - your movement will degrade the accuracy of the unit.
The documentation that comes with the system should clearly outline the required GPS settings and field procedures required for ensuring the specified accuracy of the system.
Finally, it is important that once you have collected the data you understand how to take the GPS data from a WGS-84 coordinate to a coordinate in your local coordinate system. If you are working in the sub-meter or better level of accuracy you may need to pay careful attention to the algorithm that is used for the datum transformation.
However, once you have collected the data you will want to understand the quality of the data. Tools exist for giving you precision estimates on the data. Some systems even allow you to filter out positions that do not meet your required confidence levels, so that only the good data is exported to the GIS. Some systems also allow you to export as part of the feature information metadata about the GPS position. This metadata may include the precision estimate, the DOP value and the type of correction method applied (real-time or post-processed). The Attribute Data
It is important to also ensure that the attribute data collected with a GPS / GIS data collection system is accurate and consistent with existing data. Consistent terminology will help to ensure correct interpretation of the information. This is especially important for attribute information that is subjective, such as information about the condition of a feature.
Most GPS / GIS data collection systems allow you to define the data you want to collect through the creation of a feature library or data dictionary. This data dictionary is a list of the features and their attributes that mirrors the structure of your GIS. In the field you pick the feature you are collecting and are prompted to enter the attribute information (GPS positions are logged in the background). When creating the data dictionary you can limit what is entered in the field by specifying the attribute type. Attributes can usually be specified as character, numeric, date or time, for example, if you are specifying an attribute type as numeric only numeric values can be entered. Some systems will even allow you to import the GIS table and convert it to a data dictionary. This further ensures that the data collected is consistent with existing data.
Limiting field entry appropriately will help to ensure that correct and consistent information is collected. Some systems allow further limiting of what data can be entered in the field with the creation of menu pick lists for attribute values. This has the added benefit of ensuring that the data collected in each data collection session uses consistent terminology and spelling! This is of particular importance if you have several people collecting data or are combining the data with other data. There is nothing worse than having to analyze data sets that use more than one term to describe the same thing. For example, if you have an attribute for the condition of a feature, one person might say "poor" while another might say "needs repair." Predefining the domain of an attribute will limit the terminology used.
If you want to understand further how accurate the attribute data are then you need to consider a method that allows you to cross check the attribute data collected. This will probably require the use of an alternative data collection technique. One example may be to use a digital camera image to cross check the attribute data. Today it is possible to connect a digital camera to GPS / GIS systems and take a digital image of the feature and store that as an attribute of the feature. In the office you could have someone do a random check on a certain amount of the attribute data to check that the entered data matches the attributes as seen in the digital image. Summary
So in answer to the question of "How can I ensure the data collected is accurate? And are there tools that help me to understand the accuracy of the data collected?" there are a number of ways you can do this to ensure accurate data are collected and there are tools that help you to ensure and understand the data collected. As a first step though, you need to define what level of accuracy is acceptable for the positions and attributes. Once you have defined this, it is necessary to find a GPS / GIS that is capable of achieving the specified level for positions and has the tools that allow you to collect accurate attribute data. Back |
