CMDL FY99 3rd Quarter Milestones

Predict and Assess Decadal-to-Centennial Change

Q3: Report on an intercomparison of data from the new high-resolution ultraviolet spectroradiometer recently installed at Boulder, Colorado, and a similar instrument located at Mauna Loa, Hawaii. One year of simultaneous UV spectra and ozone data will be analyzed and the relationship between UV and ozone variations will be derived. (CMDL, B. Bodhaine)

This research is related to the goal: Guide the Rehabilitation of the Ozone Layer, in the NOAA Strategic Plan element: Predict and Assess Decadal to Centennial Change. Performance measures include the use of the data for international assessments of the status of the ozone layer. Depletion of the global ozone layer has caused increases in ultraviolet radiation at the surface of the earth. However, not until recently has this correlation been clearly shown, for example at Mauna Loa Observatory where the atmosphere is free of the interfering effects of clouds and aerosols most of the time. The main goal of CMDL’s UV program is to understand the effects of atmospheric ozone changes on surface spectral UV irradiance. The measurement of UV spectra (1 nm resolution) allows the calculation of various weighted response functions (action spectra) in the UV region. The most common of these is the erythema action spectrum, defined to approximate the response of human skin to UV radiation. Another goal of the UV monitoring program is to detect possible long-term changes in UV irradiance related to any long-term changes in ozone. Because of the limited calibration accuracy of UV spectroradiometers (about 3%) it could take on the order of a decade to detect a long-term trend in UV.

UV spectroradiometers have been operating at Mauna Loa since July 1995, and at Boulder since June 1998. Clear mornings were selected for the Mauna Loa data, and clear mornings or clear afternoons were selected for the Boulder data. Total ozone data were obtained with Dobson spectrophotometers for these clear conditions at both sites. A time series was constructed for UV irradiance and ozone for each site. Changes in UV irradiance were compared with changes in ozone at various solar zenith angles. As expected, a strong negative correlation is found between UV irradiance and ozone, with a radiative amplification factor (the fractional increase in UV associated with a decrease in ozone) of about 1.1-1.2 for both sites. Because of the limited data sets at Mauna Loa and Boulder, no long-term trend can yet be detected.

Data on ozone-UV correlations are used in the WMO/UNEP Assessments of Ozone Depletion and by researchers who are developing atmospheric UV radiation models. These data are important because only a limited number of surface-based spectral UV measurements are available. Other uses of these UV data sets are to predict the consequences of changes in atmospheric ozone, and to predict the effects of these changes on plant and animal life. Finally, these data can be used by other researchers who wish to calibrate their instruments by comparison with ours.

This effort is important because it relates the UV irradiance at the surface to changes in total atmospheric ozone. Stratospheric ozone loss and its eventual recovery can have a profound effect on life on earth. Since different life forms have different UV action spectra, it is important to measure spectral UV to determine the effects of changes of UV on these different life forms. In addition, it is important to measure spectral UV irradiance and understand the effects of other atmospheric constituents on the radiation budget. For example, aerosols and clouds can have a large effect on UV and these effects are not currently well understood.

Since the project was instituted at Mauna Loa in July 1995, an almost continuous spectral UV data set has been obtained at this pristine site. This is one of the best clear sky spectral UV data sets available. The measurements in Boulder have also been successful following installation of the instrument in June 1998. Comparison of data at the two sites is difficult for a number of reasons, least of which is the altitude difference (3.4 km at Mauna Loa, 1.7 km at Boulder), causing reduced, but characterizable, UV at Boulder owing to additional molecular scattering, and the difference in latitude (19°N at Mauna Loa and 45°N at Boulder). While the Mauna Loa Observatory site generally has clear and clean skies throughout the morning hours, Boulder often clouds over several hours after sunrise, although some days may be cloud-free the entire day. Even partially cloudy skies will cause scattered UV to enter the field of view giving results which are difficult to interpret without complete characterization of the clouds. In addition, Boulder is generally subjected to some level of atmospheric aerosol (sub-micrometer sized suspended particles both of a natural haze origin and related to anthropogenic pollution) which result in absorption of UV, again in a fashion which is impossible to characterize without complete information on the aerosol particles. While these data will be compared scientifically in the future, for preliminary purposes we can compare data obtained at the two sites at the same sun angle. For the 6 months of data obtained in Boulder at CMDL’s old location in 1998, prior to the move to the new building, adequate clear sky data are available at Boulder for a solar zenith angle (SZA) of 70° (about 2.5 hr after sunrise or before sunset).

By integration of the multi-wavelength data over the erythemal spectrum one can compare, for example, the sunburning capacity of UV radiation at the two sites. Figure 1 shows the observed relationship between total column ozone, as measured by Dobson spectrophotometers, with the erythemal UV observed at a SZA of 70° on clear mornings at both sites. Data obtained at Mauna Loa are shown for the entire 1998 year by red circles. The data obtained at the same SZA on clear mornings at Boulder for the second half of 1998 when the instrument was operating are shown in blue triangles. For comparison, those Mauna Loa data also obtained during the second half of the year are also indicated by red triangles. First of all the correlation of high erythemal UV with low ozone is obvious. Second, the amount of UV is higher at Mauna Loa, for the same SZA and ozone amount, owing to the reduced atmospheric pressure, as expected. Finally, the scatter in the data is larger at Boulder, probably related to atmospheric aerosol which is nearly absent at Mauna Loa.

Figure 1. The correlation of ozone and erythemal UV for 70° solar zenith angle on clear days at Mauna Loa and Boulder during 1998. The circles show Mauna Loa data for the full year while the triangles compare Mauna Loa and Boulder data for the latter half of the year.

The two extremely low ozone (230-240 DU) - high UV (3-3.2 microwatts cm^-2) events at Boulder both occurred on November 30 (70° SZA in the morning and afternoon). Unusually low ozone values were observed at Boulder during November 1998, as can be seen in Figure 2 where the ozone and UV data are shown versus time during the latter half of 1998. This condition is believed to be related to the transport of low ozone air from low latitudes during November, which is then superimposed on already low ozone related to the depletion of ozone by halocarbons.

Figure 2. Time series of ozone (blue) and erythema (violet) on clear mornings and afternoons at Boulder during June 1998 - December 1998. Erythema data were calculated from spectral irradiance data at SZA = 70°.

The initial goals of the program have been achieved and it is anticipated that it will continue during the expected ozone recovery period during the next century.

Q3: Publish two papers on the results of measurements of the climate-forcing properties of atmospheric aerosols, one from the NOAA ground-based network and one from the NOAA P-3 research aircraft. These papers document the horizontal, vertical, and temporal variability of aerosol properties that are used in global models for calculating human effects on climate. (CMDL, J. Ogren)

The NOAA Strategic Plan element to which this research relates is to PREDICT AND ASSESS DECADAL TO CENTENNIAL CHANGE, with the objective to CHARACTERIZE THE FORCING AGENTS OF CLIMATE CHANGE. This effort contributes to the objective by characterizing the radiative properties of a major forcing agent of climate change, namely anthropogenic aerosols. The results of this research will be considered in the next IPCC assessment of the science of global climate change, which is the Performance Measure for this objective. Characterization of the radiative properties of aerosols includes a description of their horizontal, vertical, and temporal variability. NOAA's aerosol/climate monitoring program is one of the few systematic efforts underway to provide long-term, reliable determinations of aerosol radiative properties that are used in global models of human effects on climate. This particular milestone was undertaken to provide the modeling community with statistically-useful summaries of the variability of aerosol radiative properties, for subsequent comparison with model assumptions and predictions.

CMDL and its collaborators have been monitoring aerosol radiative properties at several marine and continental locations, including Sable Island, Nova Scotia; Bondville, Illinois; Cheeka Peak, Washington; Barrow, Alaska; Lamont, Oklahoma, and in central Hungary. Airborne surveys of these same properties using the NOAA P-3 research aircraft have been conducted during three campaigns and the results have been used to determine how representative the surface-based measurements are of the radiative properties of the total atmospheric aerosol distribution. This milestone involves evaluation and publication of the results from these measurements.

The primary customers are the atmospheric scientific community, in particular, modelers who calculate anthropogenic climate forcing. The results are used in assessments such as the IPCC Assessments of Climate Change. They are also of relevance to retrievals of aerosol properties and amounts from satellite-based measurements of upwelling radiance. Preliminary results have been communicated in scientific conferences and workshops. Two refereed publications are in press, another is undergoing revision following peer review, and two more are in the final stages of preparation prior to submission:

Anderson, T.L., Covert, D.S., Wheeler, J.D., Harris, J.M., Perry, K.D., Trost, B.E., Jaffe, D.J., and Ogren, J.A. 1999. Aerosol backscatter fraction and single-scattering albedo: Measured values and uncertainties at a coastal station in the Pacific Northwest. J. Geophys. Res., in press.

Sheridan, P.J., and Ogren, J.A. 1999. Vertical and regional variability of aerosol optical properties over the central and eastern United States, southeastern Canada, and the western Atlantic Ocean. J. Geophys. Res., in press

Ogren, J.A., Bergin, M.H., Charlson, R.J., Covert, D.S., Anderson, T.L., and Rood, M.J. 1999. Observations of the direct radiative forcing efficiency of continental and marine aerosols. J. Geophys. Res., submitted.

Koloutsou-Vakakis, S., Carrico, C.M., Rood, M.J., Li, Z., Shrestha, R., Ogren, J.A., Chow, J.C., and Watson, J.G. 1999. Aerosol properties and radiative forcing at an anthropogenically perturbed mid-latitude northern hemisphere continental site. J. Geophys. Res., in preparation..

Quinn, P.K., Bates, T.S., Miller, T.L., Coffman, D.J., Johnson, J.E., Harris, J.M., Ogren, J.A., Forbes, G., Anderson, T.L., Covert, D.S., and Rood, M.J. 1999. Surface submicron aerosol chemical composition: What fraction is not sulfate? J. Geophys. Res., in preparation.

The new aspect of this work is the systematic application of existing techniques to a variety of aerosol types for an extended period of time, to obtain statistically useful results. This work has also yielded a new parameter for comparing the efficiency of different aerosols for perturbing the climate. The results are significant because they provide a quantitative basis for testing the sensitivity of global climate models, as well as satellite data retrieval algorithms, to assumptions about the radiative properties of different aerosols.

As an example of the results, the figure shows the estimated error in calculated column radiative forcing resulting from lack of information on the vertical profile of aerosol radiative properties, derived from the P-3 measurements. This result suggests that aerosol radiative forcing is dominated by particles near the surface and that ground-based measurements of aerosol radiative properties at NOAA's monitoring sites can be used to estimate aerosol radiative forcing with +/- 10% uncertainties. There are indications that aerosol absorption is substantially higher during the autumn at the continental sites, possibly resulting from a seasonal cycle in agricultural practices.

Figure 3. Error in calculated column radiative forcing resulting from lack of information on the vertical profile of aerosol radiative properties for various heights of the top of the aerosol layer. This result suggests that ground-based measurements of aerosol radiative properties at NOAA's monitoring sites can be used to estimate aerosol radiative forcing with +/- 10% uncertainties.

Although the specific milestone of publishing two papers will be met as soon as the accepted papers appear in press (most likely during Q4 of 1999), the work will not stop there. In addition to seeing the other three papers through to publication, we anticipate several other papers to result from continued analysis of the data. This work has met its goal of contributing to an improved characterization of the forcing agents of climate change, but it will take longer for its impact to be felt on the IPCC assessment of the science of global climate change.