Grantee Research Project Results
Final Report: Improved Photolytic Rate Measurements at PAMS Sites
EPA Grant Number: R826772Title: Improved Photolytic Rate Measurements at PAMS Sites
Investigators: Rodgers, Michael O. , Pearson, James R.
Institution: Georgia Institute of Technology
EPA Project Officer: Hahn, Intaek
Project Period: March 22, 1999 through March 21, 2001
Project Amount: $168,930
RFA: Air Pollution Chemistry and Physics (1998) RFA Text | Recipients Lists
Research Category: Air , Safer Chemicals
Objective:
The objective of this research project was to build, calibrate, field test, and evaluate a prototype low-cost spectroradiometer suitable for accurately measuring the most important lower atmospheric ultraviolet (UV) photolytic rates at the Photochemical Assessment Monitoring Site (PAMS) sites. Furthermore, this system was designed to be sufficiently robust to operate unattended for extended periods under conditions of minimal environmental control.
Traditionally, ambient atmospheric monitoring networks for regulatory purposes have focused on observing compounds of direct interest to human health and/or welfare. Although these monitoring networks provide important information for risk assessments, they normally lack the range of measurements necessary to evaluate why these pollutants are present at the observed concentration levels.
By contrast, the PAMS network aims at providing detailed information on the chemical composition of the urban atmosphere for use in the evaluation of the accuracy of photochemical transport models and in interpretive analysis of the chemistry of the urban atmosphere, as well as more conventional regulatory purposes. Historically, a major limitation in the analysis and interpretation of data from PAMS and other similar networks has been an absence of reliable UV photochemical rates measurements at these locations. These photochemical measurements have been limited at PAMS because the necessary instrumentation has had a high acquisition and/or operational cost, or was not designed for protracted unattended operation typical of these sites.
Summary/Accomplishments (Outputs/Outcomes):
The overall program was divided into six phases. These were: (1) identification of detailed operational requirements; (2) design, construction, and bench testing of the prototype instrument; (3) development of supporting algorithms and data analysis routines; (4) laboratory and field testing of the prototype instrument at an active PAMS site; (5) comparison of PAMS photochemical rate measurements with those from other instruments and model calculations; and (6) modification, testing, and evaluation of revised instrument and/or data analysis procedures. Results from each of these phases are summarized below.
Identification of Operational Requirements
The photochemical rates selected for measurement by the prototype system were those showing the greatest importance for evaluation of fast photochemistry in urban areas. These are for ozone j(O1D), nitrogen dioxide j(NO2), hydrogen peroxide j(H2O2), nitric acid j(HNO3), nitrous acid j(HONO), formaldehyde (CH2O), methyl hydro-peroxide j(CH3OOH), methyl nitrate j(CH3ONO2), and peroxy-acetyl nitrate jPAN. Among these rates, the requirements for spectral resolution of instrument (0.3 nm) were defined by those needed to determine j(O1D), and the upper measurement wavelength limit (410 nm) was established by that necessary to determine j(NO2).
Development of Prototype Instrument
The general requirement for longterm unattended operation strongly discouraged the use of a scanning type instrument and thus only fixed grating type systems were considered. Continued improvements in both holographic UV gratings and large Charge-Coupled Device detector arrays meant that the basic bandwidth and spectral resolution requirements could be met with 2048 relatively inexpensive element arrays and high throughput single-stage spectrometers. A variety of commercial spectrometers were evaluated and a fiber-coupled 0.2-m crossed-Czerny-Turner (UV Holographic grating) system from Ocean Optics, Inc. (Model S2000) was selected as the basis for the system.
To bring the light into the system, three-inlet systems were constructed for evaluation during the field phase. These were two diffuser-type (cosine) inlets (sintered quartz and UV polymer) and a quartz ball lens. Laboratory tests confirmed that the polymer inlet gave the best overall performance as was used for most of the field trials.
Development of Data Processing Procedures
The collection of large quantities of spectra inevitably results in the collection of large volumes of data and the development of rapid and reliable data reduction techniques were essential. Accurate determination of low-light levels at high solar zenith angles (a weakness of the traditional Eppley™ TUVR UV radiometers and a prime motivating factor for the development of alternative UV measurement techniques) requires proper accounting for the impact of detector dark current on a pixel-by-pixel basis.
As detector dark current principally is a function of detector temperature, many systems provide for temperature control of the entire spectrometer system. This temperature control does, however, require additional power and provides another system failure point. Early in the development, it was decided to attempt to develop algorithms and techniques for correcting dark currents under conditions of widely varying temperatures. The rationale for this approach was that a correction algorithm effective over a wide range of temperatures could easily deal with constant temperature conditions but the converse probably would not be the case.
As part of this project, an innovative internal referencing system was developed and tested. In this approach, under nocturnal conditions system, dark currents are measured on a pixel-by-pixel basis throughout the night along with corresponding temperature data. During the day, pixels below the tropospheric UV cutoff were monitored to record changes in dark current under generally higher temperature conditions typical of daytime. These data were combined to produce correction functions for individual pixels used for daytime measurements. Field tests later verified that this approach could correct dark current to a relative error of less than 5 percent for daily temperature fluctuations of up to 13°C. The success of this approach obviated the need for mechanical shutter systems for system blanking purposes in the prototype instrument.
Data reduction, including application of system calibration factors, was explored using several approaches. The first and most obvious approach was the use of commercial spectral analysis programs (e.g., Thermo Galactic's GRAMs™ software) or instrument control programs (e.g., Labtech Notebook™ or National Instruments Labview™ were both used during the project). Although these systems worked well, their general nature added significant cost to the overall system and less expensive solutions were evaluated. These included programming dedicated routines as well as the use of spreadsheet macro programming (e.g., Microsoft Excel™). Ultimately, the final data analysis routines used for evaluation of the instrument were written as large macro programs in Microsoft Excel™ to minimize expense for potential end users. Data collection was made using the manufacturer's (Ocean Optics) data acquisition software. An example UV spectrum from the field-testing program is given below:
Figure 1. Typical UV Spectra Collected as Part of the Field Tests of the Prototype Instrument
Field Tests of the Prototype Instrument
Initial field trials of the instrument were conducted during the summer of 1999. These were generally proof-of-concept tests and were not conducted with a complete system. During Year 2 of the project (2000), the field tests of the prototype instrument were conducted at the Tucker, GA, PAMS site. The instrument was operated for various periods from April through November at this site and collected more than 150,000 UV spectra during various testing programs. The figures below show the layout of the Tucker site. The spectroradiometer is located on the top immediately to the left of the meteorological system in the left center of the picture.
Figure 2. Tucker, GA, PAMS Site Used for Field Tests
A more detailed view of the system is illustrated in Figure 3. Two reference Eppley™ TUVR UV radiometer systems are visible to the left of the main spectrometer inlet on the top of the support column on the top right of the photograph. On the top of the main enclosure between the support columns is an inexpensive "whole sky" camera system also developed as a part of this effort. This system employs a low-cost surveillance type digital camera and a hemispherical reflector. This system was used to estimate cloud coverage and type in support of modeling calculations. An illustration of a typical image from this system also is shown below. Although this system proved useful for the semi-quantitative measurements for which it was intended, it was found to lack sufficient dynamic range to be generally useful unless equipped with an automatic iris.
Figure 3. UV Radiometers and Spectrometer at Tucker, GA, PAMS Site
Figure 4. Whole Sky Image From Tucker, GA, PAMS Site
Comparison of Measurements
The final photochemical rates determined by the prototype system were in excellent agreement with a variety of multistream radiative transfer models from both the National Center for Atmospheric Research and Georgia Tech. However, achieving this good agreement required substantial post processing of the UV spectra. Automated processing of the data often resulted in sub-optimal baseline correction of the spectra, leading to artifacts such as negative photolysis rates. An example of such an artifact is shown below for an exceptionally clear day in August 2000, at the Tucker PAMS site.
Figure 5. J(O1D) for Tucker, GA, With Baseline Artifact
The presence of such artifacts resulted in substantially more data analysis activity than would be desirable for a system making routine measurements. At present, the analysis/time ratio (that is the ratio of time to analyze the data to running time) is about 0.05. That is, it takes approximately 1 hour of analyst time to process and quality assure 20 hours of data. For full-time deployment, this corresponds to about 0.2 full time equivalents of personnel per measurement site. At an annualized cost, including overhead of $100K/technican/year, this corresponds to an annual cost of $20,000 per site.
Evaluation of Instrument Development
During 2001 and parts of 2002, additional tests were performed to optimize the instrumentation and procedures developed in the earlier test years. These modifications substantially improved the reliability of the hardware but resulted in less improvement in data analysis and reduction time. The principal finding and conclusions of the research program are:
• A reliable low-cost and robust UV spectrophotometer-based system for measurement of the most important photochemical rate coefficients has been built and tested.
• The measurements from this system are in good agreement with modeling calculations under clear sky conditions and with other measurement methods.
• Data handling and data processing time remain issues. The system generates more than 50 megabytes of data per day and data cannot be conveniently transmitted over conventional telephone lines. Even weekly data pickup requires high capacity (CD-R or greater) storage media.
• The system can be deployed and maintained at low annualized costs, however, at present the cost of data analyst time ($20,000/site/year) will need to be reduced to make such photolytic measurements a routine part of the PAMS network.
• Although the inexpensive whole-sky camera proved useful for the semi-quantitative measurements for which it was intended, it was found to lack sufficient dynamic range to be generally useful unless equipped with an automatic iris.
Journal Articles:
No journal articles submitted with this report: View all 1 publications for this projectSupplemental Keywords:
Photochemical Assessment Monitoring Site, PAMS, ultraviolet, UV, environmental control, urban atmosphere, chemical composition, atmospheric optics, air, atmospheric sciences, engineering, chemistry, physics, environmental chemistry, air toxics, particulate matter, PM, tropospheric ozone, formaldehyde, nitrogen dioxide, PAMS sites, UV photolytic rate coefficients, air quality data, air quality field measurements, field measurements, hydrogen peroxide, measure, measurement methods, nitrogen dioxide (NO2), ozone, particulate matter chemistry, photolytic rate measurements, risk assessment., RFA, Scientific Discipline, Air, particulate matter, air toxics, Environmental Chemistry, tropospheric ozone, Atmospheric Sciences, Engineering, Chemistry, & Physics, Nitrogen dioxide, risk assessment, photolytic rate measurements, UV photolytic rate coefficients, stratospheric ozone, ozone, nitrogen dioxide (NO2), particulate matter chemistry, air quality data, hydrogen peroxide, air quality field measurements, field measurements, Formaldehyde, atmospheric optics, PAMS sites, measureProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.