The Pacific Exploratory Mission in the Tropical Pacific:
PEM-Tropics A, August-September, 1996

Introduction

During the early part of this decade, NASA through its Earth Sciences Program initiated the Pacific Exploratory Missions (PEM) to improve scientific understanding of human influences on the tropospheric chemistry over the Pacific Ocean. The PEM studies were conducted as part of NASA's Global Tropospheric Experiment (GTE), and, to date, have consisted of three airborne campaigns. PEM-West A & B were conducted over the northwestern Pacific during September-October 1992 (PEM-West A, Hoell et al., 1996) and February-March 1994 (PEM-West B, Hoell et al., 1997) to study Asian outflow during contrasting meteorological conditions. The third, PEM-Tropics A, was conducted during August -September 1996, and focused on the southern tropical regions of the Pacific Ocean, extending zonally across the entire Pacific Basin and meridionally from Hawaii to south of New Zealand. A fourth campaign, PEM-Tropics B, is scheduled for March/April 1999. Like PEM-West A & B, the combined PEM-Tropics A & B campaigns will provide observations during contrasting seasons (i.e., dry vs. wet). This paper describes the experimental design of the PEM-Tropics A campaign and summarizes some of the results given in the companion papers in this issue

One of the most important issues in global tropospheric chemistry is the sensitivity of the oxidizing power of the troposphere to human influence. From the perspective of global tropospheric chemistry, the Pacific basin is a very large chemical reaction vessel. From Peru to Borneo, it stretches 17,700 km in the east-west direction; the distance from the Antarctic ice shelf to Alaska is 13,300 km. The Pacific basin covers 35% of the total surface area of the earth, and 50% of the ocean surface. Since much of the Pacific basin is far removed from continental influences, observations in this region can provide sensitive indicators of the global-scale impact of human activity on the oxidizing power of the troposphere.

There also is a need to improve our understanding of atmospheric sulfur chemistry over the Pacific. Sulfate aerosols affect the Earth's radiative balance through direct back-scattering of solar radiation and indirectly as cloud condensation nuclei (CCN). CCN, themselves products of aerosol growth processes, are believed to originate through nucleation processes involving gas phase H₂SO₄ that is produced from the oxidation of SO₂ by OH. Sulfate and SO₂ over the Pacific may originate from a number of sources including long-range transport of anthropogenic pollution, marine biogenic releases of dimethlysulfide (DMS), and volcanic emissions. The relative contributions of these sources over different regions of the Pacific are still poorly known, representing a serious limitation in our ability to evaluate the role of sulfur in global climate change.

Prior to the PEM campaigns, there was little chemical data for the southern tropical Pacific region. The GAMETAG aircraft missions in 1977 and 1978 (Davis, 1980) provided some early data over the western part of the Pacific. However, these campaigns were restricted by the low ceiling and limited endurance of the aircraft, and also by the instrumentation available at the time. The more recent STRATOZ III (Drummond et al., 1988) and PEM-West A & B missions have provided detailed data along the South American and Asian rims of the Southern Pacific basin, respectively. Ozonesonde and CO measurements have been made from island sites during the SEAREX program and from ships. Even so, there were virtually no data for the southeast quadrant of the Pacific basin extending from the international dateline to the South American coast. The PEM-Tropics A observations (and those from the forthcoming PEM-Tropics B campaign) provide an extensive set of atmospheric measurements in a heretofore data sparse region. Data from this mission (e.g. airborne chemical measurements, meteorological and ozonesonde observations, and model products) along with data from all previous GTE field campaigns, have been archived in the Langley Distributed Active Archive Center (http://eosweb.larc.nasa.gov) and/or on the GTE Home Page at http://www-gte.larc.nasa.gov.

Implementation of the PEM-Tropics Campaign

The major objectives of PEM-Tropics (Phases A and B) are to provide baseline data over the southern Pacific Ocean for gases important in controlling the oxidizing power of the atmosphere, including O₃, H₂O, NO, CO and hydrocarbons, to improve scientific understanding of the factors controlling the concentrations of these gases, and to assess the resulting sensitivity of the oxidizing power of the atmosphere to anthropogenic and natural perturbations. In addition, PEM-Tropics A had three secondary objectives: (1) to survey the concentrations of aerosol precursors and ultra- fine aerosol particles over the Southern Pacific basin; (2) to improve our understanding of sulfur gas-to-particle formation over the region; and (3) to provide detailed latitude-altitude transects of long-lived gases for the evaluation of global tropospheric models.

To address these objectives, NASA, through a competitive process (Announcement of Opportunity), selected investigators who provided measurements and/or model analyses during PEM-Tropics A. The PEM-Tropics Science Team consisted of the Principal Investigator(s) from each investigative group. The Mission Scientists and Mission Meteorologists, also competitively selected, led the science team in developing the detailed design of the PEM-Tropics A campaign. Tables 1a and 1b list the DC-8 and P-3B aircraft investigations, respectively. Table 2 lists the meteorological and modeling investigations, along with the Mission Meteorologists and Scientists. The aircraft instrumentation layout is included in Figures 1a and 1b for the DC-8 and P-3B, respectively.

The NASA DC-8 aircraft has a nominal ceiling of 12 km, a cruising speed of 800 km/hr, and, as configured during PEM-Tropics A, a 10 hr flight endurance. The P-3B aircraft has a nominal ceiling of 8 km, a cruising speed of 500 km/hr, and, as configured in PEM-Tropics A, a nominal 8 hr endurance. These differing characteristics, coupled with the instrument payload of each aircraft, favored use of the DC-8 for long range transport studies and high altitude observations, and the P-3B for lower altitude, process-oriented studies. The differential absorption lidar (DIAL) aboard the DC-8 provided vertical profiles of ozone and aerosol above and below the aircraft (Browell et al., 1996), thereby enhancing transport studies conducted by the DC-8. The DIAL profiles provided a two-dimensional perspective of the structure of the troposphere in which the in situ measurements aboard the DC-8, and in some cases the P-3B, were recorded. The DIAL profiles also provided real time guidance for adjusting the DC-8 flight tracks to exploit interesting measurement opportunities encountered in-flight.

The P-3B instrument payload included measurements of a unique suite of sulfur species [DMS, SO₂, methane sulfonic acid (MSA) [gas], H₂SO₄ [gas], non-seasalt sulfate (NSS), and methane sulfonate (MS)], along with aerosol composition and size distributions, including ultrafine particles. These measurements, combined with the capability for reliable observations of OH, via the Chemical Ionization Mass Spectrometry (CIMS) technique (Eisele and Tanner, 1991), provided a unique opportunity for the focussed sulfur process studies that were conducted by the P-3B.

The broad design of the PEM-Tropics A campaign employed a series of flights from remote operational sites in the south Pacific basin. Figures 2a and 2b shows these sites, along with flight tracks of the DC-8 (fig 2a.) and the P-3B (fig. 2b) aircraft. The flight numbers in Figure 2 are keyed to Table 3a and 3b in which the major focus of each flight is summarized. As part of the overall design of PEM-Tropics, measurements obtained from the NASA DC-8 and P-3B during the intensive deployment period were augmented by ozonesonde observations from operational launch sites at Easter Island, and Lauder, New Zealand (also shown in figure 2), and by stations established by the GTE Project at Papeete, Tahiti and American Samoa. Figure 2 also shows the ozonesonde release site on Fiji which was established by the GTE project in August 1997 to support the PEM-Tropics B mission. Ozonesondes at all the sites, except Fiji, were released at a rate of one per week, beginning in August 1995, approximately one year prior to the aircraft campaign in 1996. Releases at all sites are scheduled to continue through October 1999, to encompass the upcoming PEM-Tropics B mission, and a final survey of the dry season period. During the aircraft deployment periods, the sites established by GTE and the Lauder station increased their launch rate to 2 per week. It is anticipated that the

ozonesonde data from Fiji and Samoa will provide an indicator of the seasonal changes in the gradients of trace gases that were observed to exist across the SPCZ during PEM-Tropics A (Gregory et al., this issue).

Meteorological support for real-time flight planning and post-mission analyses to assist in coupling air mass transport and chemistry is a critical element of GTE field campaigns. Meteorological support for in-field flight planning during PEM-Tropics A was provided by a Mission Meteorologist stationed with each aircraft, with supporting activities at the Massachusetts Institute of Technology (MIT) and Florida State University (FSU). Meteo France also provided significant support by allowing direct access to ECMWF gridded data and by providing valuable consultations at their Papeete, Tahiti facilities. Satellite images received and stored at MIT, along with derived products from FSU and MIT (i.e., streamline and trajectory analyses at various pressure levels, potential vorticity, etc.), were transmitted to the field operations sites at Christmas Island and Easter Island using a portable high-speed satellite-telephone data link. For operations at Christmas Island, and to a lesser extent at Easter Island, this data link provided the only access to meteorological data. A combination of the satellite data link, faxed data products, and the use of local meteorological facilities were used during operations at Papeete, Tahiti, Christchurch, New Zealand, Nadi, Fiji and Guayaquil, Ecuador.

Integration of instruments aboard the P-3B and DC-8 aircraft occurred, respectively, at the NASA Wallops Flight Facility, Wallops Island, VA, and the NASA Ames Research Center (ARC), Moffett Field, CA. The P-3B deployed approximately one week prior to the DC-8's departure, with transit flights to Christmas Island via ARC and Hawaii (see Table 3b). The original flight plans called for two local P-3B flights at Christmas Island; however, the P-3B experienced aircraft problems after the first local flight, and returned to Hawaii for repairs. After completing integration and test flights, the DC-8 departed ARC for Tahiti via Hawaii. The DC-8 and P-3B arrived in Tahiti on the same day. During this transit flight to Tahiti, both aircraft conducted substantial vertical profiling to document inter-hemispheric gradients of gases, along with a coordinated low altitude fly-by of Christmas Island to document concentrations of several key constituents measured several days earlier during the P-3B's local flight from Christmas Island.

Flight operations from Tahiti included three local flights by each aircraft, followed by transit flights by both the DC-8 and the P-3B to Easter Island. Each aircraft conducted two local flights from Easter Island after which the P-3B traveled to Guayaquil, Ecuador, and the DC-8 to Christchurch, New Zealand, with a stop over and local flight from Tahiti. While the DC-8 operated out of New Zealand, the P-3B conducted three local flights from Guayaquil to examine inter-hemispheric exchange and the primary productivity rich oceanic region near the equator prior to returning to the NASA Wallops Flight Facility. The DC-8's local flight out of Christchurch extended south to the Antarctic coast to investigate meridional gradients and the influence of transport from high latitudes on the tropical troposphere. From Christchurch the DC-8 flew to Fiji, where three local flights were conducted, followed by the return flight to ARC via Tahiti.

Overview of PEM Tropics Results

This section provides a brief synopsis of the salient results from PEM-Tropics A that are discussed in detail in the companion papers of this issue. Among its accomplishments, the PEM-Tropics A field campaign has provided a unique set of atmospheric measurements in a heretofore data sparse region; demonstrated the capability of several new or improved instruments for measuring OH, H₂SO₄, NO, and NO₂, and actinic fluxes; and conducted experiments which tested our understanding of HO_x and NO_x photochemistry, as well as for sulfur oxidation and aerosol formation processes. In addition, PEM-Tropics A documented for the first time the considerable and widespread influence of biomass burning pollution over the South Pacific, and identified the South Pacific Convergence Zone (SPCZ) as a major barrier for atmospheric transport in the southern hemisphere.

New/Improved Measurements:

The DC-8 instrument payload included a new generation photo-fragmentation two-photon laser-induced fluorescence (PF-TP-LIF) instrument for measuring NO and NO₂ (Sandholm et al., 1997; Bradshaw, et al., 1998). The improved sensitivity of this instrument provided, for the first time, accurate NO measurements at sub-pptv concentrations. Further, a high flow rate sample inlet system was employed to minimize the possibility that complex nitrogen oxides would decompose in the inlet system, which had been speculated to be a problem in previous measurements of NO₂ by this group (Crawford, et al., 1996). The PEM-Tropics A data showed that comparisons between measured NO/NO₂ concentration ratios and those computed with a photochemical steady state model agreed to within 30% (Schultz et al., this issue). This finding is in sharp contrast to results from the GTE PEM-West A mission, where the deviation between predicted and observed NO₂ was nearly a factor of 4 (Crawford et al., 1996). The agreement between modeled and measured NO/NO₂ is attributed to the improved sensitivity and the high flow rate inlet. Moreover, measurements by the PF-TP-LIF instrument revealed that NO concentrations in the marine boundary layer (MBL) were frequently below 1 pptv.

Also aboard the DC-8 was a new airborne actinic flux spectroradiometer (Shetter and Muller, this issue). This system consisted of nadir and zenith 2 pi str radiometers, and provided spectral radiance in 1 nm steps from 282 nm to 330 nm (UV-B region) and in 2 nm steps from 330 nm to 420 nm (UV-A region) at 30 seconds intervals. The accuracy of the actinic flux was stated as approximately + 11.5% in the UV-B and + 8% in the UV-A range. Shetter et al. (this issue) report that uncertainties in the resulting photolysis frequencies for O₃, NO₂, HONO, CH₂O, H₂O₂, CH₃OOH, HNO₃, PAN, CH₃NO₃, CH₃CH₂NO₃, and acetone vary between + 15% and + 20%.

Aboard the P-3B, a new airborne Chemical Ionization Mass Spectrometer (CIMS) was fielded to measure OH, H₂SO₄, and MSA. Measurements reported by Mauldin et al. (this issue a and b) throughout PEM-Tropics A, and particularly for the local P-3B flight from Christmas Island (discussed in the next section), provide convincing evidence of the capability of CIMS for airborne measurement of OH and H₂SO₄.

Sulfur Photochemisty:

One of the reasons for selecting Christmas Island as an operational site during PEM-Tropics A (and for the 1999 PEM-Tropics B campaign) is the almost ideal environmental conditions that exist in this region for an atmospheric sulfur photochemistry experiment. Specifically, Christmas Island is located where the prevailing meteorological conditions are relatively uniform with little, if any, anthropogenic contributions to the local sulfur budget. These conditions were highlighted in results from earlier ground based observations (Bandy et al., 1996) which showed a persistent anti-correlation between SO₂ and DMS over several diurnal cycles. The local flight from Christmas Island (P-3B flight 7P) was designed to exploit these conditions and the unique set of sulfur and photochemistry instruments aboard the P-3B aircraft.

The specific objective of the Christmas Island flights was to study the evolution of DMS oxidation chemistry in a common air mass from before sunrise through early afternoon. This was achieved using a Lagrangian sampling pattern within, and just above, the boundary layer. Simultaneous measurements of OH, DMS, SO₂, MSA (gas), H₂SO₄(gas), MS, NSS, and aerosol size number distribution, as well as critical meteorological parameters, permitted one of the most intensive examinations yet of the detailed chemical processes involved in the oxidation of DMS via hydroxyl radicals. Results from this flight (Davis et al., this issue) revealed distinct pre-sunrise minima in the concentrations of OH, H₂SO₄, and SO₂ that increase to a maxima at the end of the flight in the early afternoon. Concurrent with the increase in these species was a decrease in DMS beginning near sunrise. Concentrations of OH were observed to range from sunrise values near 10⁵/cm³ to noon time maximum values of 8x10⁶/cm³. Davis et al. (this issue) placed the overall efficiency for conversion of DMS to SO₂ for the conditions of the Christmas Island flight at 75%. This efficiency strongly supports the notion that SO₂ was the dominant precursor to H₂SO₄ in this marine environment. Equally, important these investigations showed that most of the SO₂ is converted to sulfate via heterogeneous processes as was found to be true also for the formation of MS. Using concurrent measurements of the key controlling species to constrain their model calculations, Davis et al. (this issue) report that the agreement between model simulations and observations for OH ranges from 5 to 20%. This suggests that for the tropical MBL, the photochemical mechanisms in current models represent those operating in the real atmosphere.

Directly related to the sulfur/photochemical study conducted on Christmas Island are the observations of new particle formation reported by Clarke et al. (this issue) during three distinctly different environmental conditions. One of these, cloud outflow, is consistent with previous observations suggesting cloud outflow as a major source for aerosol nucleation. The other two environmental conditions included (1) aged and scavenged air exhibiting characteristics of long-range transport and the influence of combustion and (2) well scavenged air within the boundary layer over productive waters high in DMS. Clarke et al. (this issue) report that all three environments exhibited similar characteristics: low aerosol surface areas, elevated sulfuric acid, and enhanced water vapor concentrations. In the case of the MBL, Clarke et. al. (1998) report results for a tropical MBL setting in which a nucleation event, forming new ultra-fine particles was for the first time directly linked to DMS oxidation via OH through the direct observation of the intermediates SO₂ and H₂SO₄(g). Clarke, et al. (this issue) note that while the nucleation observed in the marine boundary layer may have occurred under rare conditions, it nevertheless demonstrates that nucleation can occur when surface areas are sufficiently small and concentrations of water vapor and sulfuric acid are sufficiently large.

Sulfur Distribution/Sources in the Pacific:

Thornton et al. (this issue) combined measurements of SO₂ obtained during the PEM-Tropics A, the PEM-West A & B campaigns, and the Aerosol Characterization Experiment-1 (ACE-1) to produce a data set containing 4679 observations of SO₂ at altitudes ranging from 50 m to 12 km and covering a geographical region from 60 N to 72 S and 110 E to 80 W. This combined data set showed that in the northwestern Pacific, anthropogenic sources from eastern Asia dominate the sulfur chemistry in the lower troposphere out to distances of about 1500 km, and much further at mid to upper troposphere altitudes, resulting in a significant gradient between the northern and southern hemispheres. Because of the absence of significant anthropogenic sources of SO₂ in the southern hemisphere, Thornton et al. (this issue) notes that volcanic sources from east Asia may be the dominant source of SO₂ in the mid and upper troposphere of the southern hemisphere, while DMS is a significant source of SO₂ only in the tropical marine boundary layer.

Trace Gas Distributions and Biomass Burning:

An important aspect of PEM-Tropics A was to determine the extent of biomass burning influence over the remote South Pacific during the dry season of the austral tropics. The TRACE-A mission conducted in the same season had previously demonstrated a strong biomass burning influence over the South Atlantic, downwind of Brazil and southern Africa (Fishman et al. 1992). A remarkable finding of PEM-Tropics was the pervasiveness of biomass burning plumes and their impact on trace gases throughout the southern Pacific region. Flights from Fiji, New Zealand, Tahiti, Easter Island, and Guayaquil frequently encountered layers of biomass burning pollution in the 2-12 km column (Gregory et al. this issue; Schultz et. al. this issue; Talbot et. al. this issue and Fuelberg et al., this issue). Ozone mixing ratios in these layers frequently exceeded 80 ppbv and were associated with high mixing ratios of CO and other tracers of biomass burning (C₂H₂, C₂H₆, CH₃Cl, CH₃Br). Urban pollution tracers (e.g. C₂Cl₄) were not enhanced, and hydrocarbon data indicated that these pollution layers were 1-3 weeks old. The O₃/CO enhancement ratio typically was greater than 1, consistent with chemical production of O₃ and chemical decay of CO during aging. Back-trajectory analyses suggest that most of these layers originated from fires in Africa and South America and were transported to the South Pacific by strong westerly flow at subtropical latitudes (Fuelberg et. al. this issue), although analyses of AVHRR satellite images indicate that fires in Indonesia and Australia may also be sources of some of the layers (Jennifer Olson, private communication). Of particular interest is the analysis of climatological data from 1986 to 1996 by Fuelberg et. al. (this issue) indicating that the PEM-Tropics mission period was representative of the previous 11 years. Ozonesonde data for 1995-1997 at the PEM-Tropics sites, and earlier ozonesonde data at Samoa and New Zealand, also confirm that 1996 was not anomalous (S.J. Oltmans and J. A. Logan, private communication).

Analysis of acidic gases (e.g. HNO₃, HCOOH, and CH₃COOH ) measured on the DC-8 aircraft showed that over the altitude range of 2-12 km, but particularly in the 3-7 km range, air parcels were frequently encountered within 15⁰ - 65⁰ S latitude with mixing ratios up to 1200 pptv (Talbot et al. this issue). Talbot, et al. (this issue) suggest the correlation of the acidic gases with CH₃Cl, PAN, and O₃ and the absence of correlation with common industrial tracer compounds such as C₂Cl or CH₃CCl₃ indicate a photochemical and biomass burning source for the plumes. Talbot et al. (this issue) also report that the ratio of C₂H₂/CO was typically in the 0.2 - 2.2 pptv/ppbv range indicating relatively aged air masses in the plumes, consistent with the trajectory analysis reported by Fuelberg et al. (this issue).

Dibb et al. (this issue) reports on the distribution of aerosol-associated soluble ions measured aboard the NASA DC-8. The authors found low mixing ratios of all ionic species throughout the free troposphere, suggesting that the soluble ions that might have been expected in air masses influenced by biomass burning had been scavenged by precipitation. They note, however, that the activity of ⁷Be frequently exceeded 1000 fCi/m³ throughout the troposphere, indicating that the scavenging of soluble ions had occurred far up wind of the DC-8 sampling region. These observations are again consistent with the trajectory analysis presented by Fuelberg et al (this issue) and the chemical analysis by Talbot et al. (this issue) suggesting that the plumes originated from sources well west of the DC-8 sampling region. Dibb et. al. also report decreasing mixing ratios of NH⁴⁺ with increasing altitude through out the PEM-Tropics A study area, consistent with shipboard sampling indicating strong NH₃ emissions from the equatorial Pacific.

It interesting to note that observations during the P-3B flights along the west coast of South America indicated that some fresh biomass burning plumes originated from the south American continent and were transported westward by the trade winds.

An important question relative to the long-range transport of biomass burning emissions into the tropical Pacific is the impact on the general background concentrations of trace gases. Statistical analysis of the PEM-Tropics data showed that the influence of biomass burning extended beyond the plumes and pervaded the regional atmosphere, as seen for example in the strong positive correlation of ozone with CO found for the ensemble of PEM-Tropics data (see Figure 9, McNeal et al., 1998). A further indication of the level of enhancement for O₃ comes from the modeling analysis by Schultz et al (this issue) showing that advection of ozone associated with biomass burning increases the average ozone concentration throughout the troposphere by about 7-8 ppbv. Further, during both TRACE-A (Singh et al. 1996) and PEM-Tropics A (Schultz et al. this issue), biomass burning was found to be the dominant source of atmospheric PAN.

An interesting consequence of the layering associated with the biomass burning plumes observed in PEM-Tropics A was discussed by Stoller et. al. (this issue). By analyzing in situ airborne measurements obtained during PEM-Tropics A, PEM-West A & B (Newell et. al. 1996, Wu et al. 1997), and TRACE-A (Collins et al. 1996), Stoller et al. (this issue) concluded that within the troposphere "Atmospheric layers are ubiquitous", occupying approximately one fifth of the atmospheric vertical extent sampled. During PEM-Tropics A, the most commonly observed layers exhibited characteristics of air originating from biomass burning sources. The authors note that the extensive layering observed during these GTE field missions can have an important impact on atmospheric heating, and that the rates of photochemical processes in these layers will be different from those calculated using "average" atmospheric mixing ratios.

Trace Gas Distributions:

The analysis by Gregory et al, (this issue) illustrates the influence that convergence associated with the ITCZ and SPCZ have on chemical characteristics of their respective geographical regions. The authors analyzed in situ measurements obtained north and south of the ITCZ and the SPCZ, respectively. They report that in each region the air north and south of the respective convergence zones has distinctively different signatures indicative of the source regions. For example, the air north of the ITCZ exhibits a modest urban/industrial signature, while air south of the ITCZ and north of the SPCZ, is relatively clean. Consistent with other observations reported above, the chemical signature of the air south of the SPCZ was noted to be dominated by combustion products associated with biomass burning. Gregory et al. (this issue) noted that the resulting chemical gradients across each zone was more pronounced below 5 km in consistent with the strong low level convergence that is characteristic of each zone, becoming much less pronounced at higher altitudes. Back trajectory analysis by the FSU group (Fuelberg et al. this issue) also showed that much of the tropical air having low ozone mixing ratios originated east of the observation region and had not passed over land masses for at least 10 days.

Measurements of CO₂ on the DC-8 and P-3B during PEM-Tropics A resulted in the most extensive aerial CO₂ data set recorded over the South Pacific Basin. Vay et al. (this issue) analyzed PEM-Tropics A flight data combined with CO₂ surface measurements from NOAA/CMDL and NIWA to establish vertical and meridional gradients for the region. They conclude that the observed CO₂ distributions in the south tropical Pacific were noticeably affected by interhemispheric transport with northern air masses depleted of CO₂ frequently observed south of the ITCZ. Vay et al. (this issue) note that regional processes also modulated background concentrations as large scale plumes from biomass burning activities produced enhanced CO₂ mixing ratios within the lower to mid troposphere over portions of the remote Pacific. Of particular interest was a shift in the location of an apparent equatorial CO₂ source observed in the surface data between 15° N and 15° S, but realized in the airborne data between 8° N and 8.5° S, demonstrating the importance of vertical trace gas profiles in potential source/sink regions as they provide an additional constraint for global scale trace gas budget models.

Measurements of gas phase hydrogen peroxide (H₂O₂) and methylhydroperoxide (CH₃OOH) aboard the DC-8 and P-3B aircraft have also provided an extensive data set covering the region from 70⁰ S to 60⁰ N and 110⁰ E to 80⁰ W in the Pacific and 40⁰ S to 15⁰ N and 45⁰ W to 70⁰ E in the south Atlantic over an altitude range from 76 m to 13 km. O'Sullivan et al. (this issue) reports that both of these compounds exhibited a maximum concentration at a given altitude along the equator and decreasing in concentration with increasing latitude in the southern and northern hemisphere. Similar to the chemical gradients reported by Gregory et al. (this issue), O'Sullivan et al. (this issue)notes that the latitude gradient above 4 km is substantially reduced and at altitudes above 8 km there is no latitudinal dependency.

Photolysis Frequencies:

Calculated photolysis rate coefficients, i.e., J-values, are key parameters in photochemical models. In past GTE campaigns, clear-sky model calculated J-values have been adjusted for cloud effects based on differences between calculated J(NO₂) and J(NO₂) derived from Eppley radiometers. However, a continuing cause of concern has been the large systematic changes in the clear-sky baseline for the Eppley radiometers between previous campaigns (e.g., as great as a factor of 1.5). This prompted a detailed comparison of several radiometric determinations of J(NO₂) during PEM-Tropics A, discussed in detail by Crawford et al. [this issue]. As noted earlier, a new actinic flux spectroradiometer system was flown aboard the DC-8 aircraft (Shetter and Muller, this issue) to measure the actinic flux values needed to calculate J-values. Also aboard the DC-8 were zenith and nadir viewing J(NO₂) filter radiometers (see Table 1a). Aboard the P-3B, solar flux measurements were obtained using zenith and nadir viewing Eppley radiometers (see Table 1b).

The only period during PEM-Tropics A where the two aircraft were sufficiently close geographically and temporally to permit comparison of all three radiometric techniques was during flights 4D (DC-8) and 10P (P-3B) in the marine boundary layer near Christmas Island. During this period, the three radiometers exhibited trends consistent with each other and with model calculations. However, they disagreed in magnitude with the J(NO₂) filter radiometers being approximately 30 percent greater than the spectroradiometers and the Eppley radiometers falling between the J(NO₂) and the spectroradiometers. Across all DC-8 flights, agreement between the J(NO₂) filter radiometers and spectroradiometers was exceptional with regard to the variation in J(NO₂) due to clouds (i.e., R²=0.98). The J(NO₂) filter radiometers, however, continued to be consistently higher than the spectroradiometers by 30 percent. Crawford et al. (this issue) note that while model calculations agreed best with values from the spectroradiometer, the accuracy of J(NO₂) cannot be assured to better than within 30 percent.

Since each technique appears to accurately capture the variability in J(NO₂) due to clouds, the recommendation of these findings has been to normalize the variations to the clear-sky baseline of the model calculations. Although, this does not improve the accuracy of photochemical model calculations, it does ensure a consistent baseline for J-values that permits comparisons of photochemical calculations between GTE campaigns that are free of any systematic biases introduced by changes in the radiometer baselines.

A secondary finding based on spectroradiometer measurements of J(NO₂) and J(O¹D) suggests that the two J-values exhibit roughly equivalent responses to clouds. This confirms that correcting all model calculated J-values based on the cloud response of J(NO₂) should not be considered a large source of uncertainty in model calculations.

Ozone Photochemistry:

A major objective of the PEM and TRACE-A missions has been to improve understanding of O₃ production and loss in the remote troposphere. As noted above, comparisons by Shultz et al. (this issue) using the PEM-Tropics A measurements of NO and NO₂ reported from the PF-TP-LIF system reproduced the modeled ratio to within 30%, with similar success in simulating concentrations of peroxides. These results are important since they indicate that key aspects of photochemical processes in the troposphere are understood.

Model evaluation of the photochemical ozone budget over the South Pacific (Schultz et al., this issue) revealed that the tropical Pacific is a net photochemical sink. Specifically, chemical production was found to balance only half of photochemical loss. Based on the ratio (delta O₃) /(delta CO), O₃ transport from biomass burning regions was found to be significant in balancing the remaining photochemical loss. Photochemical production in the lower troposphere was also found to be largely driven by NO_x derived from the decomposition of PAN that was transported from biomass burning regions.

Photochemical model analysis of the PEM-Tropics A data has reinforced a scenario that has emerged from previous GTE missions about the factors controlling O₃ in the tropical troposphere. In this picture, which is markedly different from the prevailing view of 20 years ago, tropospheric NO_x levels are sufficiently high that O₃ concentrations in the tropical troposphere are determined by a balance between in situ photochemical production and loss. As a result, photochemical production of ozone dominates the stratospheric flux in controlling column O₃ density. Stratospheric intrusions are dramatic local events that strongly impact the local tropospheric ozone column density. However, the photochemical production of O₃ on a global scale, driven by NO_x from natural and anthropogenic sources, ultimately dominates the tropospheric O₃ budget (Schultz et al., this issue; Crawford et al., 1997a; Crawford,et al., 1997b; Davis et al., 1996; and Jacob et al., 1996). Changes in NO_x emissions in the tropics as a result of industrialization and changes in biomass burning practices would have a major impact on O₃ in the tropical troposphere and, hence, the global oxidizing power of the atmosphere.

Concluding Remarks

The PEM-Tropics A mission provided the first detailed survey of tropospheric ozone and sulfur chemistry over the South Pacific. It complemented previous GTE missions conducted in other regions of the tropics (ABLE-2 A and B, CITE-3, PEM-West A and B, TRACE-A). The observations from PEM-Tropics A showed that seasonal biomass burning in the southern tropics causes major enhancements in background concentrations of ozone and other gases over the most remote regions of the South Pacific atmosphere. From this result it appears that the biomass burning perturbation to atmospheric composition is global in scale, with major implications for the global oxidizing power of the atmosphere. Another important finding of PEM-Tropics A was that the South Pacific Convergence Zone (SPCZ) represents a major barrier to atmospheric transport, and acts in combination with the ITCZ to define boundaries for air masses over the tropical Pacific. Focused studies of sulfur chemistry conducted in PEM-Tropics A demonstrated a diurnal evolution of DMS-SO₂-H₂SO₄-OH consistent with results from photochemical models, and observed for the first time episodes of new particle formation in the marine boundary layer. Finally, the PEM-Tropics A mission featured many improvements in critical instrumentation for tropospheric chemistry including sub-pptv measurement of NO, high-quality measurements of NO₂, OH, and H₂SO₄, and spectroradiometer measurements.

The PEM-Tropics B mission to be conducted in March-April 1999 will conclude the PEM mission series by surveying the atmosphere over the South Pacific during the wet season of the southern tropics. The objectives of PEM-Tropics B extend beyond those of PEM-Tropics A to include focused studies of HOx chemistry, of the large-scale ozone minimum in the western equatorial Pacific, and of vertical transport by deep convection in the ITCZ and the SPCZ. Biomass burning influence on ozone should be near its seasonal minimum during the PEM-Tropics B period, while lightning influence should be near its seasonal maximum. Data for ozone over the South Pacific indicate particularly low concentrations in March-April [Fishman et al., 1990; Johnson et al., 1990; Oltmans and Levy, 1994]. The continuous ozonesonde network operated as part of PEM-Tropics indicates low ozone concentrations throughout the tropospheric column in March-April, with none of the high-ozone layers found in September-October. Some anthropogenic pollution could still be transported to the South Pacific during PEM-Tropics B by the circulation of Asian outflow around the Pacific High. Some interhemispheric transport of biomass burning pollution from the northern tropics is also possible. In any case, considerable contrast should be found with the conditions observed in PEM-Tropics A.

ACKNOWLEDGMENTS

The PEM -Tropics A expedition would not have been successfully conducted without the cooperation, support, and collaboration of personnel and colleagues from organizations both foreign and domestic. Many exhibited a personal interest in our endeavors and performed well beyond the normal call of duty. We give special thanks to Drs. Patrick Simon, Isabelle Leleu and colleagues at Meteo France in Tahiti. Without their dedicated efforts prior to our arrival in Tahiti, and patience during our operations in Tahiti, the PEM-Tropics A mission would not have been possible. In addition, we owe a special thanks to the DC-8 and P-3B personnel. Their dedication and patience were a critical element in translating the, often, unrealistic desires of the PEM-Tropics A science team into reality. And finally we express thanks to Mike Cadena and Fred Reisinger in the GTE contractor Project office for their dedicated support in the "care-and-feeding" of the science team during the field deployment.

References

Bandy A. R., D. C. Thornton, B. W. Blomquist, S. Chen, T. P. Wade, J. C. Ianni, G. M. Mitchell and W. Nadler, Chemistry of Dimethyl Sulfide in the Equatorial Pacific Atmosphere, Geophys. Res. Lett., 23, 741-744, 1996.

Bradshaw, J., D. Davis. J. Crawford, G. Chen, R. Shetter, M. Muller, G. Gregory, G Sashse, J. Barrick, D. Blake, B. Heikes, J. Mastromarino, and S. Sandholm, Photofragmentation to-photon laser-induced fluorescence detection of NO₂ NO: comparison of measurements with model results based on airborne observations during PEM-Tropics A , J. Geophys. Res., submitted, 1998.

Clarke, A. D., F. Eisele, K. Moore, R. Tanner, L. Mauldin, and M. Litchy, Nucleation in the equatorial environments encountered on PEM-Tropics. This issue.

Clarke, A. D., D. Davis, V. Kapustin, F. Eisele, G. Chen, I. Paluch, D. Lenchow, A. R. Bandy, D. Thornton, K. Moore, L. Mauldin, D. Tanner, M. Litchy, M. A. Carroll, J. Collins, and G. Albercook, Particle Nucleation in the Tropical Boundary Layer: A Case Study Involving Marine Sulfur Sources. In Press. Science.

Crawford, J., D. Davis, G. Chen, J. Bradshaw, S. Sandholm, G. Gregory, G. Sachse, B. Anderson, J. Collins, D. Blake, H. Singh, B. Heikes, R. Talbot, and J. Rodriquez. Photostationary state analysis of the NO₂--NO system based on airborne observations from the western and central North Pacific, J. Geophys. Res., 101, 2053-2072, 1996.

Crawford, J. H., D. Davis, G. Chen, J. Bradshaw, S. Sandholm, Y. Kondo, S. Liu, E. Browell, G. Gregory, B. Anderson, G. Sachse, J. Collins, J. Barrick, D. Blake, R. Talbot, and H. Singh, An assessment of ozone photochemistry in the extratropical western north Pacific: Impact of continental outflow during the late winter/early spring, J. Geophys. Res., 102, 28469-28488, 1997a.

Collins, J. E. Jr., B. E. Anderson, G. W. Sachse, J. D. W. Barrick, L. O. Wade. L. G. Burney, and G. F. Hill, Atmospheric Fine Structure During GTE TRACE-A: Relationships Among Ozone, Carbon, and Water Vapor. J. Geophys. Res., 101, 24,307-24,316, 1996.

Crawford, J. H., D. D. Davis, G. Chen, J. Bradshaw, S. Sandholm, Y. Kondo, J. Merrill, S. Liou, E. Browell, G. Gregory, B. Anderson, G. Sachse, J. Barrick, D. Blake, R. Talbot, and R. Pueschel, Implications of large scale shifts in tropospheric NO_x levels in the remote tropical Pacific, J. Geophys. Res., 102, 28447-28468, 1997b.

Crawford, J., D. Davis, G. Chen, R. Shetter, M. Müller, J. Barrick, M. Rodgers, and J. Pearson. An Assessment of Cloud Effects on Photolysis Rates: Comparison of Experimental and Theoretical Values, This issue.

Davis, D. D., Project GAMETAG: An Overview, J. Geophys. Res., 85, 7285-7292, 1980.

Davis, D., J. Crawford, G. Chen, W. Chameides, S. Liu, J. Bradshaw, S. Sandholm, G. Sachse, G. Gregory, B. Anderson, J. Barrick, A. Bachmeier, J. Collins, E. Browell, D. Blake, S. Rowland, D. Blake, Y. Kondo, H. Singh, R. Talbot, B. Heikes, J. Merrill, J. Rodriguez, and R. Newell, Assessment of the ozone photochemistry tendency in the western North Pacific as inferred from PEM-West A observations during the fall of 1991, J. Geophys. Res., 101, 2111-2134, 1996.

Davis, D. D., G. Chen, A. Bandy, D. Thornton, F. Eisele, L. Mauldin, D. Tanner, D. Lenschow, B. Huebert, J. Heath, A. Clarke, and D. Blake, DMS oxidation in the equatorial Pacific: Comparison of model simulations with field observations for DMS, SO₂, H₂SO₄(gas), MSA(gas), MS, and NSS. This issue.

Dibb, J.E., R.W. Talbot, E.M. Scheuer, D.R. Blake, N. J. Blake, G. L. Gregory, G. W. Sachse and D.C. Thornton, Aerosol Chemical Composition and Distribution During the Pacific Exploratry Mission, Tropics, J. Geophys. Res. PEM-Tropics this issue.

Drummond, J. W., D. H. Ehhalt, and A. Volz, Measurements of nitric oxide between 0-12km altitude and 67° N-60° S latitude obtained during STRATOZ III. J. Geophy. Res. 93, 15831-15849, 1988.

Eisele, F. L., and D. J. Tanner, Ion-assisted tropospheric OH measurements, J. Geophys. Res., 96, 9295-9308, 1991.

Fishman, J., J. M. Hoell, Jr., R. J. Bendura, R. J. McNeal, and V. W. J. H. Kirchhoff, NASA GTE TRACE-A experiment (September-October, 1992): Overview. J. Geophys. Res., 101, 23865-23880, 1996.

Fishman, J., C.E. Watson, J.C. Larsen, and J.A. Logan, Distribution of tropospheric ozone determined from satellite data, J. Geophys. Res. this issue, 95, 3599-3617, 1990.

Fuelberg, H. E., R. E. Newell, S. Longmore, Y. Zhu, and D. J. Westberg, and E. V. Browell. A meteorological overview of the PEM-Tropics experiment. This issue.

Gregory, G. L., D. J. Westberg, M. C. Shipham, D. R. Blake, R. E. Newell, R. W. Talbot, B. G. Heikes, G. W. Sachse, B. A. Anderson, and D. C. Thornton. Chemical characteristics of Pacific tropospheric air in the region of the ITCZ and SPCZ. This issue.

Hoell, J. M., D. D. Davis, S. C. Liu, R. Newell, M. Shipham, H. Akimoto, R. J. McNeal, R. J. Bendura, and J. W. Drewry, Pacific Exploratory Mission-West A (PEM-West A): September-October 1991. J. Geophys. Res., Vol. 101, No.D1, Jan. 20, 1996, pp. 1641-1653.

Hoell, J. M., D. D. Davis, S. C. Liu, R. E. Newell, H. Akimoto, R. J. McNeal, and R. J. Bendura, Pacific Exploratory Mission-West B (PEM-West B): February-March, 1994. J. Geophys. Res., Vol. 102, No.D23, Dec. 20, 1997, pp. 28,223 - 28,239.

Jacob, D. J., B. G. Heikes, S.-M. Fan, J. A. Logan, D. L. Mauzerall, J. D. Bradshaw, H. B. Singh, G. L. Gregory, R.W. Talbot, D. R. Blake, and G. W. Sachse, Origin of ozone and NO_x in the tropical troposphere: A photochemical analysis of aircraft observations over the South Atlantic Basin, J. Geophys. Res., 101, 24,235-24,350, 1996.

Johnson, J.E., R.H. Gammon, J. Larsen, T.S. Bates, S.J. Oltmans, and J.C. Farmer, Ozone in the marine boundery layer over the Pacific and Indian Oceans: Latitudinal gradients and diurnal cycles, J. Geophys. Res. this issue, 95, 11,847-11,856, 1990.

McNeal, R.J., D.J. Jacob, D.D. Davis and S.C. Liu, The NASA Global Tropospheric Experiment: Recent Accomplishments and Future Plans, IGACtivities NewsLetter, Issue No. 13, 2-18, 1998.

Mauldin, R. L. III, D. J. Tanner, and F. L. Eisele, Measurements of OH During PEM-Tropics, This issue a.

Mauldin, L., D. III, D. J. Tanner, J. A. Heath, B. J. Huebert, and F. L. Eisele, Observations of H₂SO₄ and MSA during PEM-Tropics A, This issue b.

Newell, R. E. , Z.-X. Wu, Y. Zhu, W. Hu, E. V. Browell, G. L. Gregory, G. W. Sachse, J. E. Collins jr., K.. K. Kelly, and S. C. Liu, Vertical Fine-Scale Atmospheric Structure Measured From the NASA DC-8 During PEM-West A. J. Geophys. Res., 92, 2071-2081, 1987.

O'Sullivan, D. W., B. G. Heikes, M. Lee, W. Chang, G. Gregory, D. Blake, and G. Sachse, The distribution of hydrogen peroxide and methylhydroperoxide in the Pacific and South Atlantic Troposphere , This issue.

Schultz, M., D. J. Jacob, J. A. Logan, Y. Wang, D. Blake, J. Bradshaw, G. Gregory, G. Sachse, R. Shetter, H. Singh, and R. Talbot, On the origin of tropospheric ozone and NO_x over the tropical Pacific. This issue.

Sandholm, S., S. Smyth, R. Bai, and J. Bradshaw, Recent and future improvements in two-photon laser-induced fluorescence NO measurement, J. Geophys. Res., 102, 28,651-28,662, 1997.

Singh, H., D. Herlth, R. Kolyer, L. Salas, J. D. Bradshaw, S. T. Sandholm D. D. Davis, J. Crawford, Y. Kondo, M. Koike, R. Talbot, G. L. Gregory, G. W. Sachse, E. Browell, D. R. Blake, F. S. Rowland, R. Newell, J. Merrill, B. Heikes, S. C. Liu, P. J. Crutzen, M. Kanakidou., Reactive nitrogen and ozone over the western Pacific: Distribution, partitioning and sources, J. Geophys. Res., 101, 1793-1808, 1996.

Shetter, R.E. and Müller, M., Photolysis frequency measurements on the NASA DC-8 during the PEM-Tropics Mission using actinic flux spectroradiometry: Instrumentation description and results, J. Geophys. Res. PEM-Tropics this issue, 1997.

Stoller, P., R. E. Newell, M. A. Carroll, B. E. Anderson, J. D. W. Barrick, E. V. Browell, G. L. Gregory, G. W. Sachse, S. Vay, J. D. Bradshaw, and S. Sandholm, Measurements of atmospheric layers from the NASA DC-8 and P-3B aircraft during PEM-Tropics A. This issue.

Talbot, R. W., J. E. Dibb, E. M. Scheuer, D. R. Blake, N. J. Blake, G. L. Gregory, G. W. Sachse, J. D. Bradshaw, S. T. Sandholm, and H. B. Singh, Influence of biomass combustion emissions on the distribution of acidic trace gases over the southern Pacific basin during austral springtime. This issue.

Thornton, D. C., A. Bandy, A. Driedger, B. Bloomquist, and T. Wade, Sulfur dioxide distributions over the Pacific Ocean 1991-1996. This issue.

Vay, S. A., B. E. Anderson, G. W. Sachse, J. E. Collins, Jr., T. J. Conway, D. R. Blake, D. J. Westberg, L. O. Wade, T. A. Slate, and G. D. Nowicki, Airborne Observations of the Tropospheric CO₂ Distribution and its Controlling Factors over the South Pacific Basin. This issue

Wu, Z.-X., R. E. Newell, Y. Zhu, B. E. Anderson, E. V. Browell, G. L. Gregory, G. W. Sachse, and J. E. Collins, Jr., Atmospheric layers measured from the NASA DC-8 during PEM-West B and comparison with PEM-West A, J. Geophysical Res., 102, 28,353-28,365, 1997.