How well can people observe the flow state of temporary streams?

Even though more than half of the global river network does not have continuous flow, temporary (i.e., non-perennial) streams are poorly represented in traditional monitoring networks. Therefore, new approaches are needed to monitor these streams. Citizen science provides an interesting opportunity as people, equipped with smartphones, can observe the flow state of temporary streams. Such observations can go beyond a simple classification of flow vs. no flow and include ecologically important states, such as standing water, isolated pools, or wet streambeds. However, the quality of citizen science data for temporary streams has so far not been thoroughly assessed. Therefore, we asked more than 1,200 people during 23 field days to visually determine the flow state of eight temporary streams based on six classes ranging from a dry streambed to flowing water. Participants could most clearly distinguish a flowing stream from a non-flowing stream. The overall agreement between participants was 66%; 83% of the selected flow states were within one class of the most frequently selected flow state. The agreement with the expert was lower (56% chose the same class, and 79% chose a state within one class). Inconsistencies between the selected flow state and answers to specific yes-no statements about the temporary stream were largest for the dry streambed and damp/wet streambed states. These discrepancies were partly caused by participants looking at different parts of the stream (i.e., participants considered the flow state for a location further upstream or downstream). To ensure that all participants determine the flow state comparably, we recommend clear definitions of the flow state classes, detailed information on the exact location for which the flow state needs to be determined, as well as more training.