I would like to take another shot at a question asked during the ICWP webinar “Securing More Funding for Water Monitoring”. In reference to the fact that ICWP was able to successfully lobby for a USD 7.2 million increase in budget for the USGS, it was observed that, at the scale of the United States, that doesn’t seem like much money. How much of an investment would be required to make a significant difference in events like the California drought or the current Texas flooding?
It is true the increase only amounts to about USD 1.30 per km2 when distributed over the US land mass. That amount would be a modest increase in budget for any scale of water monitoring program. However the question is much more interesting for considering whether the return on investment in water monitoring has an upper limit.
Is benefit/cost fully scalable?
The California drought is estimated to be costing USD 3 billion per year. California needs to be careful what it wishes for. Texas has ended its devastating 5 year long drought with massive flooding costing several lives and millions of dollars in property damage. There is risk of being unprepared for extremes at both ends of the hydrological spectrum. What has happened in Texas, and what is happening in California, could happen anywhere, anytime.
Being adequately prepared can substantially reduce costs.
Some of the costs are a result of water over-use. Highly compelling hydrological information is needed to stimulate the political will to codify, rectify, or replace archaic and arcane water legislation. Some costs are due to water misuse. Trusted hydrological information is needed to inform policies that can effectively prevent water misuse and provide adaptive management guidance during periods of extremes. Some costs are a result of water abuse. Continuous, real-time, hydrologic information is needed to monitor compliance with, and enforce infractions of, beneficial water policies and regulations.
Some of the costs are due to failures in floodplain planning and management. Floodplains are in high demand for use as transportation corridors and the development of homes and industrial uses. The choice is whether to allow development and limit risk by construction of expensive flood works (e.g. dams, levees, floodways) or to prohibit development at a substantial cost to the local economy. High quality hydrological information is required to evaluate, and continuously validate and update, the assumptions about flood dynamics upon which costs and benefits depend.
Some of the costs are due to failure in the engineering design of stream crossings, water storages, spillway designs, and flood works. High quality hydrological information is needed to inform engineering design. The information must be representative of the hydrological characteristics at the scale of the project being designed, this requires a high density monitoring network. This information must be able to identify if the hydrological regime is stable through time and to correctly characterize any change that is occurring, whether due to land-use change, climate-change, or any other reason. Designs based on hydrologic specifications that are no longer valid need to be re-designed. This requires stream gauges with a long period of record.
Some of the costs are due to hydrological models that either under, or over, predict the most impactful events resulting in either a failure to respond, or an overly-disruptive response, to events as they occur. Better model design, calibration, validation, and initialization is only possible with better hydrological information provided by improving monitoring.
I could go on, but let us suppose the incremental benefit is 10 times the incremental cost for water monitoring.
The data and information resulting from water monitoring can substantially improve outcomes for impactful hydrological events that occur at many time and space scales. Can this information make a substantial difference for the largest scale events?
Let us suppose, as a thought experiment, that the ICWP was 100 time more successful. Instead of USD 7.2 million, they got USD 720 million in increased funding for the USGS. Would this be a waste of money, or would our assumption of benefit/cost be valid if the amount of funding is increased by 2 orders of magnitude?
To start with, the quantity and quality of data would likely scale in a non-linear way to the increase in funding. The increase in station density would reduce the mean distance between stream gauges increasing the proportion of time spent at a gauge versus on the road. The increased density would greatly improve the opportunity for automated cross-location inter-comparison reducing the time spent reviewing data. Increased funding would result in faster evaluation, and implementation of new and emerging technologies resulting in further efficiencies.
The investment in new and emerging technologies would likely result in rapid and substantial advances in the objective quantification of data uncertainty. This would have a profound effect, not only on data interpretation and analysis but also in field methods and procedures. As the impact of various sources of uncertainty are measured and monitored, field operations can be optimized to reduce, or eliminate, the most addressable sources of uncertainty.
One hundred times more funding would not only increase station density it would increase the ability to monitor for detection of change as previously discontinued stations with long record are re-activated. A long data gap at a reference gauge is not ideal but it is better than not having any long records in a landscape that has been impacted by human influence.
How much impact would such a funding scenario have on decisions that are made at all relevant time-scales? There is great resistance to change many of the legal, social, and economic factors that exacerbate risk. Is it possible that the weight of new evidence would be sufficient to overcome this inertia? People have short memories — is there some threshold scale of information that needs to be crossed in order for people to understand, and respond to, their vulnerability to hydrological events?
Can the day-to-day benefits of better water accounting justify the expense of monitoring for infrequent, high-payback, events?
Given that a single, highly impactful future decision (e.g. design of a water storage reservoir) can justify decades of monitoring long before the need was known, is anyone betting against the need for highly impactful decisions in almost every watershed within the next 50 to 100 years? How much funding would be required to ensure that all future decisions are well-informed and evidence based? What do you think? Please comment.
If you missed it and would like to watch a recording of the webinar, click here.