The four most important elements of an observation are a photo, a date, the coordinates (i.e., latitude and longitude), and a community-supported identification. When an observation has these four elements, it is considered “research grade”. However, there is a fifth element that is nearly as important as these other elements in determining the value of an observation to researchers, but it often gets overlooked. This fifth element is accuracy, which is the proximity of the measured latitude and longitude to the actual location where the observation was made. The accuracy can be determined by a GPS device or manually by the user using a mapping app, such as the iNaturalist map function.

The importance of having observations with high accuracy cannot be overstated. Only a small number of uses of iNaturalist observations can accept observations with low accuracy. For example, obtaining a species list for Griffith Park, Balboa Park, or one of the ten counties in southern CA, does not require observations with high accuracy. However, the vast majority of uses require high accuracy, which I am roughly defining as within 50 meters. For example, one of the key, immediate uses of RASCals observations is to understand how species distributions have responded to urbanization. Analyses will ask questions like, how are urban distributions correlated with lot size, age of the neighborhood, or distance from the nearest creek or urban park? Thus, small offsets of only 100 or 200m could dramatically alter the conclusions by moving observations onto house lots of different size, age, or distance to the nearest creek.

Thankfully, it is easy to assign accuracy with iNaturalist. The easiest approach is to use the iNaturalist smartphone app because the app will assign accuracy automatically. A good strategy here is to watch the accuracy value (labeled “Acc.”) in the “Location” section of the “Add observation” page. This is the page that automatically opens on the smartphone app when you click to add an observation. Once you open the app, the accuracy value will start dropping as your location gets pinpointed with greater accuracy. You can also see the accuracy improving by clicking on the map page in the app and watching as your phone’s GPS device narrows in on your location. The accuracy is depicted as a circle around your location; as this circle gets smaller, the accuracy is improving. Once the accuracy value has stopped decreasing, you can save the observation. Importantly, you do not need to wait for your GPS device or smartphone to pinpoint your locality before you take the photo through the iNaturalist app. As long as you stay stationary before hitting the save button, your phone will narrow in on your location and report a high accuracy for your observation.

If you are using the metadata from a photo to record the location, or using the map feature on iNaturalist.org, then you will need to manually add the accuracy value. (Note that accuracy values are typically not stored in the photo metadata.) When editing the observation on iNaturalist.org, click “edit” in the locality section, and the point will show up on the map. If there is not already a value listed for accuracy, enter a value and a red circle will show up around your point. You can drag that circle in and out until you have set it to an appropriate accuracy value.

Critically, the goal here is not to get the accuracy value as small as possible. The goal is to have a value that reflects you or your GPS device’s best assessment of accuracy. For example, if you are uploading a photo that lacks latitude and longitude data, and all you can remember is that you took the photograph while on a hike in Griffith Park, then the accuracy value should be several kilometers to accommodate the entire area in which you could have made the observation. This allows future users of the observation to appropriately assess your observation and decide whether or not it is appropriate to include in their analyses.

Here are some observations with appropriate accuracy values:

(http://www.inaturalist.org/observations/756344) – this observation has a small accuracy value, because I could tell that my phone’s GPS mapped the location to the exact piece of exposed concrete where I made the observation.

(http://www.inaturalist.org/observations/420731) – this observation has a large accuracy value. I took the photo with a digital camera lacking GPS capabilities while on a mtn. bike ride. When I submitted the photo, I could not be certain of the exact road cut where I made the observation. As a result, I placed the point at the most likely spot but gave a large value for accuracy to cover nearby road cuts where the observation could also have been made.

If you have questions about this post, please comment below or message me or Richard Smart through iNaturalist.

Greg Pauly: http://www.inaturalist.org/people/gregpauly
Richard Smart: http://www.inaturalist.org/people/rsmart