Monetizing the Value of Science Missions

Scientific missions in a budget-constrained environment have difficulty showing a monetary value of knowledge accretion. In the case of some Earth observations however, such a demonstration is ‘low hanging fruit'. For example, suppose a mission (or suite of missions) promises to yield decisive information on the rate of global warming within a shortened time frame. How much should we (mankind) be willing to pay for this knowledge today? The US interagency memo on the social cost of carbon (US‒SCC) creates a standard yardstick for valuing damages from carbon emissions . It also creates the framework for value of Information (VOI) calculations placing a dollar value on uncertainty reduction. Without passing judgment on the validity of the assumptions underlying the USMSCC, we illustrate how these VOI calculations can be used to monetize the relative value of different Earth Observing Systems. We follow the US-SCC, stipulating uncertainty only on climate sensitivity with a truncated Roe Baker distribution, stipulating discount rates of 2.5%, 3% and 5%, using one of the Integrated Assessment Models sanctioned in US-SCC (the 2009 Excel version of DICE , Nordhaus 2008) and in basing social cost of carbon only the climate damages (the economic benefits accruing from increased emissions are omitted).

The Business as Usual (BAU), DICE Optimal, and Stern emissions scenarios provide a wide range of mitigation policies. Relative to BAU, the modest DICE Optimal scenario reduces carbon emissions a factor of 2 by 2065 and is carbon neutral by 2115, while the stronger Stern scenario reduces carbon emissions a factor of 7 by 2065 and is carbon neutral by 2085. If the parameters are not uncertain, the optimal policy is computed by finding a savings rate and abatement policy such that the value of an additional unit of abatement equals the value of an additional unit of emissions. Under uncertainty, the optimization becomes burdensome (we cannot simply optimize sample-wise). The DICE Optimum policy is optimal if climate sensitivity (cs) is equal to 3.

To illustrate results, suppose that we are on the BAU emissions scenario, and that we would switch to the Stern emissions path if we learn with 90% confidence that the decadal rate of temperature change reaches or exceeds 0.2 [C/10yr]. Under the US‒SCC assumptions, the year in which this happens, if it happens, depends on the uncertain climate sensitivity and on the emissions path. The year in which we become 90% certain that it happens depends, in addition, on our Earth observations, their accuracy, and their completeness. The resolving power of a climate observing system cannot exceed the natural variability in the climate system. All climate observations add noise to the natural variability arising from a variety of observing limitations. Key examples include calibration errors and orbital sampling uncertainty.

When two climate observations differ in their resolving power, the relative value of the more powerful system can be monetized using the US‒SCC. If the more powerful climate observation detects a 0.2C/10yr rate of increase 10 years earlier than the less powerful system, then by switching to the Stern emissions path based on the more powerful system, we avoid damages caused by these 10 years of BAU emissions. The economic value of the averted damages depends on the discount rate, and the years in which the damages occur. A new climate observation would be economically justified if the net present value (NPV) of the difference in averted damages, relative to the existing systems, exceeds the NPV of the system costs. We present illustrative results comparing the proposed CLARREO advance in satellite absolute calibration for climate change records to an existing system for detecting decadal temperature change and cloud feedback. Cloud feedback is the key uncertainty driving a factor of 3 uncertainty in climate sensitivity (IPCC, 2007). In this example, we therefore examine both the advances possible for infrared observed temperatures as well as reflected solar observations of cloud radiative effects.

In addition to the above mentioned parameters, the VOI depends on the required confidence level, the trigger value at which we would abandon the BAU emissions path, the path to which we switch, and the date at which the new system is launched. The VOI of CLARREO in this decision context is the surfeit of NPV of averted damages, relative to the existing system. Over all it is in the order of tens of trillions of US dollars. Among the noteworthy conclusions are (1) whether the emissions path to which we switch is the DICE optimal or the Stern path makes only a modest difference in the VOI of CLARREO, (2) raising the trigger value from 0.2C/10yr to 0.3C/10yr, increases he VOI of CLARREO, while at the same time increasing the total NPV of climate damages, and (3) the choice of discount rate affects the VOI by a factor 5 ~ 6.

While the example is given in the context of CLARREO advances in satellite absolute calibration accuracy, societal decisions are made only when a large number of rigorous observing systems are consistent. As a result, the VOI can be considered to be relevant for improvements in the full climate observing system, including confirming observations of a wide range of climate variables including carbon cycle, ice sheets, sea level, cloud properties, aerosols, and many others. Since the NPV costs of typical climate observing systems are orders of magnitude lower than the value of the information they provide in a wide variety of decision strategies, the value of investing in enhanced climate observing systems, based on the US interagency memo on the social cost of carbon, is overwhelmingly evident. This also argues for continual enhancement of the SCC assessment process.