And yet the two questions associated with this session still lack clear answers: How much data do we really need to do our jobs? And who should or should not pay for those data to be produced? Arguably the answer to both questions is "it depends." As unsatisfying and unhelpful as that answer might seem to be, the author contends that this overarching answer can be further refined to deliver a rational and repeatable analysis method that can be used to examine situations one-by-one and come up with consistent and fair answers to both questions - how much data is enough, and who pays?
For the first question of "how much data is enough?", the more expansive answer is "it depends on what decisions you need to make with the data". For instance, if you're making decisions about where to route air traffic around severe weather, synoptic-to-mesoscale data is likely sufficient. Additional data is always welcome, but if completely avoiding that severe weather is the operational objective, then having stations every 25 to 40 km is likely good enough, depending on the geography and the type of weather involved.
On the other hand, if a farmer is trying to figure out the optimum schedule for watering and pesticide application on a field-by-field basis, synoptic and mesoscale data is not going to be sufficient. In such cases, higher resolution data is needed and the federal government is not going provide data on the needed resolution, so the farmer faces a business decision: how much would it cost to provide my own data at a high enough resolution to matter, and will I recoup more than that expense in increased profits earned because of the increased productivity that the data enables?
In fact - this analysis that the farmer is doing provides the key- there is a two-way analysis being done: how much does the effort cost, and how much economic return will result from that effort? Government procurement systems focus almost exclusively on the former and far more rarely on the latter - but given finite and limited resources, both questions need to be answered to make a sound decision. Not only do both questions get looked at in private sector decisions, but a further bar must be cleared - investors are investing their own personal funds and so are motivated to cast a tight eye on the project to ensure that the net benefit and return on investment will be positive. If it doesn't, the project loses the investor's money - something that serves to fine tune and tighten the decision making (and reduce the need for government investment and risk-taking).
Given the additional financial return analysis and related requirements applied to private sector decision making, private sector investment in sensing should at least be considered in every case and should be selected in any case where it makes sense, even if that role has traditionally been fulfilled by the federal government. As proposed in multiple forums in recent years, this will likely lead to a framework in which the federal government provides certain foundational observing capabilities and inherently governmental functions, while the private sector fills in the gaps with sensors and data tailored to specific business problems. The boundary between those solution sets has been and always will be constantly shifting with new technology and new business problems, but with active coordination and cooperation, the two sectors can work to complement each other to provide a better consolidated level of service to the Weather Enterprise and the nation.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner