its really eye opening to see so many professionals discussing uncertainties that they faced, im quite new at this but i do have a project in mind and i think that the uncertainty that would be faced is the spatial uncertainty, as the place is planned to undergo improvements there would be changes around the area.
I have a slightly different thought about uncertainty;
The world we work in is in essence unpredictable - there may be general patterns we can recognise under certain conditions, however, if conditions change the patterns may be distrupted. I think a really imprtant point about uncertainty is that it is inevitable (even if we could accurately know the state of a thing the information would be out of date by the time we did anything with the data). So, what are the ways in which we can work well with uncertainty? Be really clear about why, what and how we measure (and really understand and be transparent about the limitations - donât over predict or generalise) and understand itterative (try and learn) approaches to project management and delivery (expect things to change and adapt).
I imagine representing the boundaries between villages and the size of each village (spatial) will be a good example of uncertainty.
⌠and also village names!. In many countries, a single village can go under different names depending on the communities referring to it. Eg. local languages, national language, international language or whether the name referes to the area in which the village is in or the settlement itself. Its a common problem for international humanitarian work where collation of critical information at times of disasters is critical - but if data is collected with reference to different named locations, it can be very complicated to get a true picture of the impacts of a disaster. A common approach to this is the use of âP-Codesâ (place codes) which give a unique hierarchical reference to each settlement location, often supported by multiple attribute fields to record the different names it also goes by.
1 Like
Being certain of the uncertainty is always a good approach 
Satellite imagery can be temporal uncertainty. Where the snapshots of the field data donât necessarily reflect the current data as some faulty may have occurred along the way and uncertainty results may occur. Uncertainty is not mainly caused by inability to measure but it can results in mutated or faulty happening during recording which can be caused by nature such as wind. The field data is produced to simplify what is happening around the world. Uncertainty data may delay the process of discovering what is happening around the world or what is detected on the satellite imagery as the results would not be clear. Uncertainty is inevitable but it can be reduced and lessen the chances of it occurring frequently, firstly by using more effective techniques to collect accurate data which is expensive and by fixing the uncertain data
Iâm a student and new to this field. I n imagine some of the uncertainty would lie in the interpretation of the images/data. I imagine uncertainty is inevitable. We are inundated with data so I assume there would be a certain degree of simplification required to make sense of all this data.One way to reduce the source of uncertainty would be to physically examine the area, if possible.
Part of uncertainty can come from harsh weather that disturbing me to collect data set of number of animals in the nature reserve. It can be inevitable, because I cannot stop it. It destruct my data collection due to uncontrollable weather. Weather it is something that is fluctuating and nothing i can do to reduce it.
I do not have any GIS work to share experience as this software is new to me, so have nothing to share on uncertainty, but it is great to learn from others.
when using camera traps and collar tracking, there is spatial uncertainty because it is difficult to map out precise boundaries of cheetah territories so we deliberately simplify the data just to have an informed idea of their movement patterns.
I encounter temporal uncertainty since seagrasses have seasonal growth, with extent and quality of cover changing within 1 year or across years. I also see temporal uncertainty in catch data, depending on season and across years.
ah, an âunknown unknownâ as opposed to a âknown unknownâ if were aware of examples of uncertainty - phraseology courtesy of Donald Rumsfeld (âThere are known knowns, known unknowns and unknown unknownsâ to paraphrase).
Thank you for this input!
For my example, weâre going to determine wolf population size. In addition to researchers, we usually engage a larger number of volunteers (students, hunters, guides, âŚ) in our project. We all look for suitable samples in the field (usually feces), and thatâs how we get the locations of the animals. If I focus only on these patterns and in this way of obtaining data, there is, of course, a great deal of uncertainty. Many volunteers are not yet fully skilled in recognizing wolf droppings, or the feces is old and therefore confusion occurs with the feces of other animal species. With such a huge number of participants, there are also errors in entering GPS coordinates. Due to space and time constraints, there is also uncertainty in the data. To keep this to a minimum, such censuses take a very long time, even several years. This includes other ways of estimating population size (e.g. camera traps, GPS collars, DNA analysis, monitoring wolves by provoking howling, âŚ).
1 Like
The uncertainty that exists in my field data is Temporal Uncertainty.
There is always some degree of uncertainty when working with data sets. Satellites can be affected by weather on Earth and solar flares. When working with satellite tagging, we may know where an animal is at certain times of the day or when they âpingâ but we can only connect the dots in between, we donât actually know where those animals were. Making sure to get data with high resolution will help reduce uncertainty.
Satellite pictures might be affected by weather conditions eg clody day hence some aspects on the ground might not be captured or even over emphasized hence can be misleading. data with high magnification will definitely be of great significance to reduce uncertainties.
Well in the data that I take from birds there is mainly temporal uncertainty due to the fact that the richness of birds will vary according to the time of sampling, the weather and the season. These uncertainties could be avoided by increasing the sampling frequency, however this would require more budget and resources
I am also going to have to try to figure out the uncertainty of the visual data that I plan to work on.
As noted in the description, visual data need to be simplified because of the complexity of real-world data. When we generalize the heterogeneous geospatial dataset, the result is degraded resolution, limiting temporal and spatial details and sensitivity.
I think the uncertainty in GIS data representation varies from field to field and the type of information and analysis to focus on. For example, in the case of watershed analysis, the average condition of the spatial feature of the landscape will have to be taken into consideration, losing the detailed information of the exact location and the condition of the individual streams and water bodies. Some of the information that can be lost is the direct/indirect runoff, condition of the incised channels, quality of buffer zone, etc. This situation can make planning and management recommendations tricky.
As others have also pointed out, these uncertainties can be reduced by frequent field visits to compensate for the lack of information and careful layering of different geospatial datasets.
Thinking back on a bird population census I was a part of there were many different uncertainties. The accuracy of the gps was a known uncertainty as we knew how many metres it was off by. There was differing abilities of observers to spot birds and record the bands accurately, this was managed by working in teams with and having at least 1 experienced person in each team. A big known uncertainty would be that we would not have seen every bird. The census is conducted over one week once every 5 years so provides only a short snapshot of information that would limit the ability to accurately model population trends and where birds are located.