1

Final Report, February 8, 2018 Partially Satisfying Task 6

WORK ORDER UNDER THE CONTRACT BETWEEN TCEQ AND CONTRACTOR: The University of Texas at Austin

GAD Number 582-13-30089-012

Tracking No.: 2015-5107-PCR#57139

Biogenic VOC Monitoring Prepared by Dave Sullivan, Ph.D. The University of Texas at Austin, Center for Energy and Environmental Resources Building 133, MC R7100, 10100 Burnet Rd. Austin, TX 78758-4445 Phone: 512-471-7805. Email: sullivan231@mail.utexas.edu

Contents 1. Background ................................................................................................................................. 3

1.1 Monitoring Instruments ........................................................................................................ 3

1.2 Big Thicket Radio Tower Site .............................................................................................. 3

2. Summary of Major Activities ..................................................................................................... 7

2.1 Schedule of deliverables ....................................................................................................... 7

2.2 Site Search and Selection ...................................................................................................... 7

2.3 Site Installation and Initial Operation ................................................................................... 8

2.4 Collocation Experiment ...................................................................................................... 12

2.5 Site Shutdown ..................................................................................................................... 12

2.6 Tree Survey ......................................................................................................................... 13

3. Data Validation ......................................................................................................................... 14

3.1 Isoprene ............................................................................................................................... 14

3.2 Solar radiation and photosynthetically active radiation ...................................................... 18

3.3 Winds and temperature measurements ............................................................................... 25

4. Scientific Approach .................................................................................................................. 35

4.1 Experimental design............................................................................................................ 35

5. Data Analyses ........................................................................................................................... 39

5.1 Ambient Isoprene and Other Data ...................................................................................... 39

5.2 Colocation experiment ........................................................................................................ 44

2

5.3 Meteorological Data Collection .......................................................................................... 49

6. Conclusions and Recommendations ......................................................................................... 50

Appendix 1 Isoprene Analysis in Ambient Air SOP (by Orsat LLC) .......................................... 51

A1.1 Description: ...................................................................................................................... 51

A1.2 Equipment: ....................................................................................................................... 51

A1.3 System Settings: ............................................................................................................... 52

A1.4 Quality Control: ................................................................................................................ 52

A1.5 Operator Responsibilities: ................................................................................................ 52

A1.6 System Maintenance:........................................................................................................ 53

A1.7 System Calibration: .......................................................................................................... 53

A1.8 References: ....................................................................................................................... 53

A1.9 Airgas certification ............................................................................................................. 55

Appendix 2 Tree Survey ............................................................................................................... 56

Tree counts in 28 sampled 0.1 acre locations ........................................................................... 56

Extrapolation ............................................................................................................................. 59

Parker Forestry University of Texas Isoprene Study Submitted Report: Big Thicket National Preserve January 2018 .............................................................................................................. 64

3

1. Background The Texas Commission on Environmental Quality (TCEQ) and other organizations have found that biogenic emission computer models over-predict isoprene (C5H8) concentrations compared with measurements from urban surface Automated Gas Chromatograph (auto-GC) monitors. TCEQ's existing network of auto-GCs are in urban areas or locations without significant isoprene emitters (oak forests). Accurate information as to isoprene concentrations is important in accurately computer-modeling ozone concentrations. In conducting this research project, The University of Texas at Austin (UT) has provided hourly, or higher frequency, biogenic isoprene and meteorological observations at a monitoring site for two summer ozone seasons. The monitoring location had been sited where TCEQ's biogenic emission model predicts elevated concentrations, which will help evaluate the performance of the biogenic and photochemical computer model results. A summary of major activities from July 2015 to February 2018 is shown in Section 2. Section 3 describes the data validation. The original project description appears in Section 4. A summary of the collected isoprene data appears in Section 5. Section 6 is short summary of findings. One appendix contains a detailed description of the isoprene instrument and how it operates, and a second appendix describes a tree survey conducted around the monitoring site in December 2017 and January 2018.

1.1 Monitoring Instruments UT was responsible for site operation and maintenance and data collection for the instruments listed in Table 1. Table 1 Measurements taken for this project

Instruments Parameter Time resolution

Planned period of operation

Peak Performer 1 Reducing Compound Photometer (RCP) 910-132 isoprene instrument

Isoprene (C5H8), parts per billion carbon (ppbC)

5-minutes & 1-hour

Mar. 2016 – Oct. 2016, Mar. 2017 – Oct. 2017

Meteorology Climatronics 460 Wind speed mph, wind direction degrees, ambient temperature, degrees F

5-minute & 1-hour

Mar. 2016 – Oct. 2016, Mar. 2017 – Oct. 2017

Photosynthetically active radiation (PAR) instrument

Spectral range from 400 to 700 nm, in µmol (photons) m-2 s-1

5-minute & 1-hour

Mar. 2016 – Oct. 2016, Mar. 2017 – Oct. 2017

Solar radiation pyranometer Solar energy in the visible range in watts/m2

5-minute & 1-hour

Mar. 2016 – Oct. 2016, Mar. 2017 – Oct. 2017

1.2 Big Thicket Radio Tower Site This remote monitoring site at 30.5445, -94.3455 has electricity and unimproved road access. Figure 1 shows a topographic map of the area and Figure 2 and Figure 3 show Google Earth Pro aerial images of the site and the surrounding area. The site is 5 miles south of FM 1943 off the Bumstead Cemetery Rd (CR 4800), or 3 miles east of Hwy 69 on the Triple D Ranch Rd (CR 4780). Much, if not all, of the county roads in the area are unpaved. Figure 4 is a photo of the monitoring station and fenced pad site. Since the end of the project the site has been restored to

4

its original condition, with exception of reseeding the top soil, which was left exposed for natural regrowth at the direction of the Big Thicket staff biologist. Figure 1 Topographic and land cover map of the area around the monitoring site

5

Figure 2 Location of Big Thicket isoprene monitoring site, 30.5446N, 94.3465W

Figure 3 Site is 34 miles NNW of Beaumont, a few miles (on unimproved roads) from the NPS HQ

6

Figure 4 Big Thicket Radio Tower isoprene monitoring station and pad site

7

2. Summary of Major Activities

2.1 Schedule of deliverables Table 2 summarizes major activities in the project and their associated deliverables sent to the TCEQ. Table 2 Project deliverables and schedule

Deliverable Deliverable Due Date Grant Activities Description (GAD) (Task 1) Deliverable 1.1: TCEQ approved GAD Deliverable 1.2: TCEQ approved Quality Assurance Project Plan (QAPP)

July 24, 2015 July 24, 2015

Monitoring site evaluation (Task 2) Deliverable 2.1: Monitoring Site Proposal Report

September 1, 2015.

Isoprene and Meteorological Monitoring (Task 3) Deliverable 3.1: Monitoring Operations Report and Preliminary Data.

By the 15th of each month (or the following workday if on weekend or holiday).starting October 2015 Monitoring March 15, 2016 – October 15, 2016, and March 15, 2017 – October 15, 2017

Data Validation (Task 4) Deliverable 4.1: Validated data

Within 60 days of the end of monitoring from Task 3

Final Report (Task 5) Deliverable 5.1: Draft Final Report Deliverable 5.2: Final Report

Draft January 25, 2018 Final February 8, 2018

Two additional sub-tasks had been added for 2017. During the summer of 2017, project operations were interrupted so the isoprene instrument could be moved twice to be installed in two TCEQ monitoring stations with Perkin Elmer TD100 automated gas chromatographs (auto-GCs) that measure hourly isoprene concentrations along with 45 other hydrocarbon species. This intercomparison was part of the quality assurance for the instrument. Secondly, a tree survey was conducted in the area around the Big Thicket Radio Tower site during December 2017 and January 2018. In addition to reports to the TCEQ, the research permit to use the Big Thicket required that UT submit annual reports and a final report to the National Park Service Integrated Resource Management Applications (IRMA) data portal. UT submitted all three reports, with the 2017 report also submitted as the final report.

2.2 Site Search and Selection UT evaluated the historical isoprene measurements at auto-GC monitoring sites around Southeast Texas, as well as modeling results showing predicted isoprene concentrations around the region. UT conducted a one-day driving trip around southeast Texas to scout for potential isoprene monitoring locations on September 3, 2015. This was described in Monitoring Site

8

Evaluation Report, September 1, 2015 Rev. 1, September 8, 2015. Existing monitoring sites were visited, as were potential site locations in the Trinity River Wildlife Refuge and the Big Thicket National Forest. Later, the selection of the radio tower clearing within the Big Thicket National Forest was made jointly by TCEQ, UT, and the U.S. National Park Service (NPS) staff. Before proceeding with installation at the Big Thicket site, and after discussions with the NPS staff, UT developed a testing procedure for communications at the site to ensure that wireless communication would not interfere with the other wireless operations nearby, and that the other operations would not affect the project’s operations. Testing was successful. Figure 5 shows an image for the ground selected for installation of the temporary monitoring station for this project. Figure 5 Grounds west of the radio tower before installation of the CAMS

2.3 Site Installation and Initial Operation The trailer, the meteorological and solar radiation equipment, and the isoprene analyzer were moved to the site, and the analyzer operated in a warm-up and test mode on February 24, 2016, and began measuring ambient air on March 4, 2016. The trailer initially was parked adjacent to the shed housing electronic equipment for the nearby radio tower, as work was delayed on site construction when NPS notified UT that an archeological survey of the site would be required. The archeological survey of the site was conducted for Thursday March 17, 2016. The summary email from Mr. Jarvis sent on March 19 is as follows:

9

I completed 19 shovel tests, each to a depth of 50 cm (about 20 in.), at the proposed isoprene monitoring station in the Big Thicket on Thursday, 17-March. No cultural materials were present in any of the tests. Based on the negative results of shovel testing and the SHPO’s determination that no historic properties will be affected, I recommend clearance for the project to proceed. I will compile a more formal letter report detailing the results of the survey sometime in the next couple of weeks. Please feel free to forward this message to any interested parties and do not hesitate to contact me with any questions or concerns. Best regards, Jonathan Jonathan H. Jarvis, MLA, M.S., RPA Associate Director, Texas Archeological Research Lab Lecturer, Department of Anthropology The University of Texas at Austin Phone: 512/471-5959 www.utexas.edu/cola/tarl www.texasbeyondhistory.net

Subsequent to the successful archeological survey, the NPS granted a construction permit for the pad site on March 22, 2016. In early April 2016, the pad site and fence were installed, and arrangements were made to install a power pole in mid-April. Photos of the development of the site appear in Figures 6 through 10. Per the agreement with the NPS, the top soil that was removed from the 30 by 30 feet square pad was piled in a berm and covered so that when monitoring ended and the gravel pad was removed, the site could be quickly restored to baseline conditions. The first year of operations continued through mid-October (October 17, 2016). Over the winter the isoprene analyzer was stored offsite. It was reinstalled in March 2017, and good data were being received by March 21, 2017. On June 14, 2017, a lightning strike to the nearby radio tower caused serious damage to the project Sutron data logger. This happened despite the normal precautions taken in site set-up, which includes a lightning grounding rod. The data logger was shipped back to Sutron and no replacement was available. After the instrument was repaired and returned, the project switched to the collocation experiment and the isoprene analyzer and peripheral equipment were moved to the Nederland High School auto-GC site.

10

Figure 6 Crew on site to begin pad construction April 7, 2016

Figure 7 Pad in place, top soil in berm, looking west, April 7, 2016

11

Figure 8 Pad site, fence in place, looking west, April 12, 2016

Figure 9 Pad site, fence in place, looking south, April 12, 2016

12

Figure 10 Pad site, fence in place, looking east, April 12, 2016

2.4 Collocation Experiment The project equipment for measuring isoprene was installed at the Nederland High School TCEQ monitoring site and began collecting data on July 31, 2017. The instrument ran at Nederland until August 14, at which point it was moved to the Lake Jackson TCEQ monitoring site. Data were collected at Lake Jackson from August 16 until August 23, at which point the TCEQ began shutting down monitoring equipment in preparation for Hurricane Harvey and associated severe weather. Results of the collocation experiment are included in a later section of this report. After the area-wide conditions improved following the Hurricane, the isoprene analyzer and peripherals were restored at the Big Thicket site and ran from September 12 to October 25, 2017.

2.5 Site Shutdown At the conclusion of the isoprene measurements, UT and its contractor secured all the monitoring equipment and UT returned the equipment, the trailer, and meteorological tower to Austin. UT disconnected power and communications and UT contracted with a local company to remove the fencing, power pole, and pad, and restore the topsoil. The research permit had originally called for mulching and reseeding of the pad site. However, when UT inquired as to as to the materials and grass type in early January 2018, the following reply was received:

Thu 1/11/2018 3:04 PM Don't worry about mulching or seeding the area, we can let the lawn grass take over on its own. The lawn is Bahia grass which is non-native, and I can't really advocate planting an exotic species!

13

Andrew Andrew J. Bennett Biologist, Big Thicket National Preserve 6044 FM 420, Kountze, TX 77625

Figure 11 shows the condition of the surface left after the pad and fence removal. Figure 11 Former pad site surface left after shutdown

2.6 Tree Survey UT investigated how to do a tree survey around the monitoring site. The idea was to create an estimate of the actual ground conditions to compare with satellite and other land use estimates that are used in biogenic emission models. In addition, the survey addressed data validation purposes of the isoprene measurements, in that it confirmed the presence of known isoprene emitters surrounding the monitoring site. UT consulted with two researchers from the Texas Forest Service and contacted several potential certified foresters in commercial practice who provided tree survey services. The resulting tree survey was earlier provided to the TCEQ and is included as an appendix in this Final Report.

14

3. Data Validation Over the course of the study, the five-minute time scale data were studied for data quality. This included the isoprene, wind speed, wind direction, temperature, solar radiation, and photosynthetically active radiation data. Date were then generally averaged by hour for reporting to the TCEQ.

3.1 Isoprene The model for data validation for the isoprene data was TCEQ SOP DATA-024, which is written for the Perkin Elmer Ozone Precursor automated gas chromatograph (auto-GC). The contractor for this project also validates data from the auto-GCs operated by UT and is experienced in the procedures. Step to data validation are summarized below. As part of the Peak instrument operation, calibration curves were run at the initial start of sampling and at the end of sampling each year. Additionally calibrations were performed if any significant maintenance was performed. Quality Control samples were used to ensure that the data produced from the analyzer were of a known quality, and consistent throughout the analytical process. Nightly quality control checks along with analytical blanks were run to insure that the response of the instrument remained consistent and did not vary. Quality control limits were such that corrections were required if the nightly standard check was outside of 25% in difference, or if the zero-air blank recorded a 2 ppbC concentration or higher. The overall agreement for step points with the calibration curve had to be < 10% relative standard deviation. The contractor investigated and corrected problems resulting any failure of quality control data to meet the quality objectives. The contractor performed daily remote checks to review daily calibrations. The contractor spot checked the chromatograms to ensure no interferences and proper integration on a weekly basis. The contractor visited the site once a week to check gas bottles and the dilution system for pressures. Additional site visits were conducted depending on the results of routine site checks and quality control sample results. The contractor flagged all data associated with failure of quality control as invalid. A second level of data validation was performed by the UT Principal Investigator, when large volumes of data were available. At this level, data were examined as 5-minute time scale data were averaged into one-hour values. The UT PI qualitatively assessed where an hour with fewer than nine 5-minute time scale samples could represent a valid one-hour. In regulatory context, general 75 percent (9 of 12) of data must be present. In this scientific study, hourly averages were kept if the data appeared to agree with surrounding hours and similar hours on nearby dates. The UT PI also assessed the likelihood that unusually low or high concentrations were associated with local meteorological conditions. The contractor and UT PI collaborated on a large scale colocation experiment in which the Peak instrument was moved and operated collocated with two TCEQ auto-GCs, as is described later in this report. This showed from a wider view that the Peak instrument correctly identified and quantified isoprene concentrations. Figures 12, 13, and 14 show time series for the isoprene 5-minute values that have been censured from the data for quality purposes in that it is very unlikely than ambient isoprene was measured at the reported concentrations.

15

Figure 12 Apparently bad isoprene value on March 17, 2016

Figure 13 Suspected bad isoprene values on March 28, 2016

16

Figure 14 Suspected bad isoprene values on October 13, 2016

Figure 15, on the other hand, shows an example of a period of elevated isoprene for which there was no evidence of spurious or false data. On August 11, 2016, the highest short term (and the highest one-hour average) values were measured. Figure 16 shows that solar radiation was relatively high on August 11, and Figure 17 shows that measured wind speeds dropped around the time of peak concentration, suggesting stagnation. Together these two factors present an assignable cause for the elevated isoprene on August 11, 2016. Figure 15 Potentially valid elevated isoprene values on August 11, 2016

17

Figure 16 Data from Figure 15 as hourly mean isoprene , solar radiation around August 11, 2016

18

Figure 17 Data from Figure 15 as hourly mean isoprene, resultant wind speed around August 11

3.2 Solar radiation and photosynthetically active radiation This project included measuring solar radiation and photosynthetically active radiation (PAR). Measurement of solar radiation at UT’s Big Thicket was made with a pyranometer and recorded in watts per square meter units. Measurement of photosynthetically active radiation (PAR) at UT’s Big Thicket site was made with a sensor that measures light in the 400-700 nm waveband recorded in micromoles per square meter – second (μmol m-2 s-1). These units are based on the number of photons in a certain waveband incident per unit time (s) on a unit area (m2) divided by the Avogadro constant (6.022x1023mol-1). This is also referred to as Photosynthetic Photon Flux Density (PPFD). Below is an extract of explanatory text from the Apogee website http://www.apogeeinstruments.com/conversion-ppf-to-watts/ (accessed January 2018)

Conversion - PPF to Watts (for UV light in sunlight) *19 October 2016: UV conversion factor changed from 0.342 to 0.327 based on measured solar spectra for clear skies at multiple zenith angles and overcast skies. This new multiplier is 4.4 % lower than the old multiplier, which was based on the energy content of photons at 350 nm. The new multiplier was derived from spectral data, but corresponds to the energy content of photons at 366 nm. We recommend back correcting data when possible.

To convert units, multiply the measurement from the sensor, in units of µmol m-2 s-1, by 0.327 J µmol-1. This converts the UV measurement to units of J m-2 s-1, which is W m-2.

19

This multiplier is only suitable for sunlight measurements. If the UV sensor is used to measure different radiation sources (for example, electric lights), a different multiplier is required for unit conversion because the multiplier is dependent on the spectrum being measured. A spectral measurement is required to derive the conversion factor.

In general the ratio of PAR (μmol m-2 s-1) to solar radiation (W m-2) is around 2.0.1 In this work the PAR has been converted to W m-2 units for direct comparison to the pyranometer results. This would predict a ratio of 0.654. For the March to October 2016 measurement period the midday ratio (slope of linear regression) was 0.67 and for the shorter March to June, Sept. to Oct. 2017 measurement period was 0.61. The difference in measurement periods does not appear to be a factor, in that a month to month comparison between the two years shows the 2016 slopes were higher than in 2017. The last two months of measurements in 2016 were the lowest slopes by month in 2016 and were closer to the average for 2017 months. The graph showing the comparison of all hourly PAR to solar radiation caused some concern for data quality. This graph appears in Figure 18, in which PAR is presented in its original μmol m-2 s-1 units. The slope is very close to 2.0, but, there appear to be two linear relations present, with the fitted regression line in between. As mentioned above, this is from a different observed slope in 2016 compared to 2017. Figure 18 also shows many points for which there is not good linear agreement (meaning a consistent ratio) between the two parameters. The data were examined on an hour by hour and month by month basis. The month to month variations were small, but the time of day assessment revealed information. Figures 19 and 20 show relationship for PAR vs Solar Radiation in W/m2 for 2016 and 2017, respectively, for the hours between 5 and 9 CST inclusive. In these two figures, the darker highlighted points that are away from the straight line are from several different days at or near 8 CST. It is possible that shadows from the surrounding heavily forested area may be an explanation. Figures 21 and 22 show the regressions for PAR vs Solar Radiation in W/m2 for 2016 and 2017 for the midday peak solar intensity hours between 10 and 14 CST inclusive. Figure 21 for 2016 shows a slope of 0.674 with R2 > 99%, and Figure 22 for 2017 shows a slope of 0.610 with R2 > 99%. Figure 23 shows for months for which a month to month comparison is possible, the mid-day slopes. This only involves May, June, September, and October. For May and June the 2017 slope is statistically significantly lower, while the slopes are lower in 2017 but not statistically significant for September and October. Despite the variations in the slopes, the actual differences may not be practically important. Finally, Figures 24 and 25 show the show relationship for PAR vs Solar Radiation in W/m2 for 2016 and 2017, respectively, for the hours between 15 and 19 CST inclusive. In these two figures, the darker highlighted points that are away from the straight line are from several different days at or near 16 CST. It is possible that shadows from the surrounding heavily forested area may be an explanation.

1 Jacovides, C.P. et al., “Ratio of PAR to broadband solar radiation measured in Cyprus” Agricultural and Forest Meteorology 121, Issues 3–4, 20 February 2004, 135-140 https://doi.org/10.1016/j.agrformet.2003.10.001 found at https://www.sciencedirect.com/science/article/pii/S0168192303002508?via%3Dihub Jan. 2018

20

For all other hours – 0 to 4 CST and 20 – 23 CST – the recorded values above 0.0 for PAR and solar radiation were infrequent, as the sun was either below the horizon or too low in the sky. Figure 18 PAR (μmol m-2 s-1) vs solar radiation (Watts/m2) at Big Thicket, all data

21

Figure 19 Over the months of operation in 2016, for the hours between 5 and 9 CST inclusive, the relationship for PAR vs Solar Radiation in W/m2, highlighted points all clustered close to 8 CST

Figure 20 Over the months of operation in 2017, for the hours between 5 and 9 CST inclusive, the relationship for PAR vs Solar Radiation in W/m2, highlighted points all clustered close to 8 CST

22

Figure 21 Over the months of operation in 2016, for the hours between 10 and 14 CST inclusive, the relationship for PAR vs Solar Radiation in W/m2

23

Figure 22 Over the months of operation in 2017, for the hours between 10 and 14 CST inclusive, the relationship for PAR vs Solar Radiation in W/m2

24

Figure 23 Comparison of mid-day 10 – 14 CST inclusive slope PAR vs solar 2016 vs 2017

Figure 24 Over the months of operation in 2016, for the hours between 15 and 19 CST inclusive, the relationship for PAR vs Solar Radiation in W/m2

25

Figure 25 Over the months of operation in 2017, for the hours between 15 and 19 CST inclusive, the relationship for PAR vs Solar Radiation in W/m2

3.3 Winds and temperature measurements Task 4 for this project required quality assurance steps to meet TCEQ Monitoring Division Standard Operating Procedure (SOP) DQRP-016 for meteorology with appropriate modifications for vertical wind measurements. As part of the monitoring done in the Big Thicket National Forest for this project, a set of Climatronics wind and temperature measuring equipment were operated at the site. The electric service at the site was not as good as previously experienced in more urban and small town settings, and some damage from lightning storms – despite precautions taken – reduced some of the met data collection. This section describes the data validation for the wind speed, wind direction, and temperature data collected. Wind speed and direction data measured at the UT trailer data were deemed to be of less importance in the project than other parameters for the following reasons.

1. Because of the close proximity of the trailer to trees, the wind data were often, if not always, somewhat compromised, and

2. Because wind and temperature data were simultaneously collected at the radio tower under the University of Utah MesoWest program’s WRRT2 Southern Rough site. MesoWest data can be downloaded from there at http://mesowest.utah.edu/cgi-bin/droman/download_api2.cgi?stn=WRRT2.

The University of Utah provided UT with several references to the quality of the MesoWest data. The response message from U. of Utah follows:

Thank you for contacting MesoWest. The data MesoWest collects from site WRRT2 is real-time provisional data provided to us by the Bureau of Land Management Wildland Fire Management Information System. Our group is in the process of developing and improving upon real-time and retrospective data check procedures that can be used to help isolate potentially incorrect data

26

(additional discussion here https://synopticlabs.org/api/mesonet/reference/qc/) but these data have not undergone extensive quality control review and checks beyond these initial procedures. Additional information on data usages can also be found here: https://synopticlabs.org/api/legal/. This station is a part of the RAWS Fire Weather Station Network, and there is some more detailed documentation here on the standards utilized for these stations (https://famit.nwcg.gov/sites/default/files/pms426-3.pdf). For questions specific to this site, there is a RAWS Help Desk contact email which may be helpful (rawshelp@blm.gov). The station is called Southern Rough RAWS and has a satellite identifier of FA634432, which may be useful if you reach out to the Help Desk. Regards Alexander Jacques, Postdoctoral Research Associate MesoWest/SynopticLabs and Atmospheric Sciences University of Utah, NSCC Rm. 484 (801) 581-4362, alexander.jacques@utah.edu

The WRRT2 monitor is recorded as sitting at 30.5447, -94.3461, which places it 44 meters east of the UT site, at a location within the clearing which would still likely to be affected by nearby trees under some conditions. Despite the potential effects of the adjacent forest, the surface winds for this project have been assessed as described below. Using the WRRT2 data for comparison, the wind speed and direction data were assessed for data quality per the procedures in DQRP-016 in which data are compared to other monitoring sites. In addition, UT used NOAA Air Resources Laboratory gridded model-derived datasets from the National Weather Service's National Centers for Environmental Prediction Eta Data Assimilation System (EDAS)2. These datasets are used in the HYSPLIT trajectory modeling program. UT ran one hour back-trajectories from 10 meter starting height above ground level (AGL) at the Big Thicket monitoring site for each hour from March 1 to October 25, 2016 and from March 1 to October 2017. The resulting 11,472 one hour trajectories, which were represented by the start and end latitudes and longitudes and start and end altitudes (meters AGL) were used to calculate one hour resultant wind speeds and directions, and vertical wind speed. Data from the three sources are shown graphically below:

1. UT’s station measurements 2. WRRT2 station measurements 3. Derived HYSPLIT estimates

Figures 26, 27, and 28 shows the statistical distributions for UT wind direction (wd_UT), wind speed (ws_UT), and ambient temperature (temp_UT). The wind speed statistics from UT reflect the effects of the surrounding tree canopy. The mean wind speed for 5,071 hourly observations was only 1.7 miles per hour (mph), and only 10 percent of observations were 3.6 mph or greater. Nevertheless, the wind direction distribution shows the expected highest frequency winds associated with south-southeast directions. There are fewer valid wind direction values (3,642) because wind directions for wind speeds less than 0.5 mph were treated as invalid and set to missing values.

2 See ftp://arlftp.arlhq.noaa.gov/pub/archives/edas40/readme.html (accessed February 2018)

27

Figures 29, 30, and 31 show the statistical distributions for WRRT2 wind direction (wd_WR), wind speed (ws_WR), and ambient temperature (temp_WR) for March 1 through October 31, 2016 and March 1 through October 31, 2017. The wind direction distributions are fairly similar between the UT and WRRT2 sites, but there are fewer very light wind speeds at WRRT2 and overall a mean wind speed of 2.0 mph, with 10 percent of observations 5.0 mph or greater. Note that wind speeds at WRRT2 are reported as integer values. There are a nearly complete set of hourly over the two years with 11,528 observations for wind speed and temperature, but only 6,772 observation of direction, because very low wind speeds were not mapped to a reported direction. Figures 32 and 33 show the statistical distributions for the derived HYSPLIT wind direction (wd_HY), wind speed (ws_HY). Like the UT and WRRT2 data, the HYSPLIT directions have a strong south-southeast mode, but the data are spread out more. The mean derived wind speed is 4.3 mph, with 10 percent of winds greater than 8.3 miles per hour. In this comparison, data from Hurricane Harvey on August 30 and 31, 2017, were not used. Figure 34 shows the distribution for the derived vertical wind speeds, presented in meter per second (mps) units, which are very small numerically. Figures 35 and 36 are regressions of UT hourly and derived HYSPLIT wind speeds to nearly coincident WRRT2 wind speeds. Note that the time tag for the WRRT2 data is 38 minutes past the hour. In this analysis, the 0:38 CST WRRT2 value was compared to the 0 CST UT and HYSPLIT values, 1:38 CST compared to 1 CST, etc. The HYSPLIT and WRRT2 speeds are very highly correlated with a slope near 1.0 (0.91) but a 2.5 mph intercept explaining the higher speeds from HYSPLIT. The UT and WRRT2 speeds are also highly correlated, but the UT wind speeds are significantly lower overall than the WRRT2 speeds. Figure 37 shows a very close match between the hourly temperature UT measurements and nearly coincident WRRT2 temperature measurements. Figure 38 and 39 show comparisons between the HYSPLIT and UT hourly wind directions to nearly coincident WRRT2 wind directions. For this assessment, only HYSPLIT wind directions greater than 45 and less than 315 degrees were used (to avoid the wrap-around at 360/0 degrees) and only recorded or derived wind speeds greater than 2.5 mph were used. The results show that HYPLIT derived directions match well with the measured directions at the two sites. All three data sets are provided along with this final report. UT recommends using the HYSPLIT derived data as being more representative of well exposed air movement.

28

Figure 26 Statistical distribution for hourly resultant wind direction at UT Big Thicket site

Figure 27 Statistical distribution for hourly resultant wind speeds at UT Big Thicket site

29

Figure 28 Statistical distribution for hourly ambient temperatures at UT Big Thicket site

Figure 29 Statistical distribution for hourly wind direction at MesoWest WRRT2 Big Thicket site

30

Figure 30 Statistical distribution for hourly wind speed at MesoWest WRRT2 Big Thicket site

Figure 31 Statistical distribution for hourly ambient temperature at MesoWest WRRT2 Big Thicket site

31

Figure 32 Statistical distribution for hourly derived wind direction using HYSPLIT at the Big Thicket site

Figure 33 Statistical distribution for hourly derived wind speed using HYSPLIT at the Big Thicket site

32

Figure 34 Statistical distribution for hourly derived vertical wind speed using HYSPLIT at the Big Thicket site, meters per second units

Figure 35 Regression comparison of UT hourly wind speed to nearly coincident WRRT2 wind speed

33

Figure 36 Regression comparison of HYSPLIT derived wind speed to nearly coincident WRRT2 wind speed

Figure 37 Regression comparison of UT hourly temperature to nearly coincident WRRT2 temperature

34

Figure 38 Regression comparison of HYSPLIT hourly wind direction to nearly coincident WRRT2 wind direction

Figure 39 Regression comparison of HYSPLIT hourly wind direction to coincident UT wind direction

35

4. Scientific Approach

4.1 Experimental design From the Proposed Grant Application and the Grant Application Description come the following listing of five tasks for the project. Task 1: UT will compose the Grant Activities Description and Quality Assurance Project Plan. Task 2: UT will evaluate TCEQ's network of CAMS sites in eastern Texas (east of 99° West Longitude) to determine if an existing CAMS location is sited where elevated biogenic isoprene concentrations are expected. In particular, UT will study data from 34 currently active automated gas chromatographs (auto-GCs) that measure hourly isoprene concentrations. An example of the method for evaluating existing sites is shown in Table 3. Table 3 shows the mean concentration of isoprene in parts per billion-carbon (ppbC) units from the TCEQ’s Danciger site (Brazoria County, 29.149 N, 95.765 W) averaged over a 10 year period (2005 – 2015) and summarized by month and hour (Central Standard Time (CST)). By comparing all sites in this manner, the suitability of existing sites may be judged. Figure 40 shows a graph of the Table 3 data that provides another means to compare data from existing sites. In addition, UT will use existing forest, land use, land cover, and biome data sources as well as biogenic emission modeling output to choose potential monitoring sites. An example of modeling output is shown in Figure 41. Information from various other studies of biogenic emissions will also be considered. Several research papers that are being reviewed are listed in the Reference section report. Access to the site via developed roads; the availability of utilities, including internet; space for the monitoring trailer; and prevailing wind conditions will be considered. A proposal detailing at least three potential monitoring locations (including site photographs, utility connections, biogenic modeling isoprene emission estimates, forest type, and surrounding anthropogenic emission sources) will be delivered to the TCEQ project manager by September 1, 2015. UT expects that the site location will be chosen by the TCEQ Project Manager by September 15, 2015.

36

Table 3 Mean isoprene ppbC from Danciger, 2005 – 2015, by month and hour

CST Jan. Feb. Mar. Apr. May Jun. Jul. Aug. Sep. Oct. Nov. Dec. 0 0.02 0.03 0.04 0.12 0.22 0.66 0.52 0.57 0.40 0.25 0.10 0.04 1 0.02 0.03 0.04 0.12 0.20 0.53 0.45 0.39 0.36 0.22 0.08 0.04 2 0.03 0.02 0.04 0.14 0.22 0.36 0.35 0.49 0.31 0.20 0.09 0.04 3 0.03 0.02 0.04 0.13 0.21 0.35 0.33 0.46 0.22 0.18 0.09 0.04 4 0.02 0.03 0.04 0.12 0.20 0.35 0.38 0.39 0.23 0.16 0.08 0.04 5 0.02 0.03 0.05 0.12 0.19 0.38 0.40 0.39 0.21 0.14 0.08 0.04 6 0.02 0.03 0.05 0.22 1.04 3.38 2.68 1.38 0.42 0.16 0.07 0.04 7 0.02 0.03 0.08 0.71 2.52 6.75 6.91 7.04 3.62 1.20 0.16 0.05 8 0.03 0.05 0.13 0.97 2.66 5.24 5.57 6.43 4.93 2.58 0.52 0.08 9 0.05 0.05 0.16 1.10 3.07 5.19 4.72 5.01 4.31 2.19 0.60 0.11

10 0.05 0.06 0.24 1.47 3.76 5.72 5.23 5.58 4.28 2.03 0.52 0.11 11 0.05 0.06 0.35 1.81 4.34 6.59 6.45 6.19 4.54 2.19 0.55 0.11 12 0.06 0.08 0.45 2.08 4.66 6.91 6.88 6.77 4.95 2.27 0.59 0.12 13 0.06 0.09 0.52 2.14 4.74 7.64 7.53 7.48 5.17 2.37 0.61 0.13 14 0.06 0.09 0.50 2.04 4.78 8.02 7.60 8.15 5.23 2.40 0.59 0.13 15 0.05 0.08 0.44 1.80 4.54 7.97 8.18 8.24 5.29 2.33 0.52 0.11 16 0.04 0.07 0.36 1.52 4.17 7.82 8.46 8.48 5.05 2.40 0.48 0.10 17 0.04 0.06 0.26 1.27 3.44 7.28 7.88 8.54 5.24 2.80 0.59 0.11 18 0.03 0.04 0.17 1.09 2.73 6.53 7.13 8.18 4.15 1.87 0.37 0.09 19 0.03 0.03 0.09 0.66 1.51 4.09 4.63 4.58 2.20 1.11 0.25 0.06 20 0.03 0.02 0.06 0.44 0.81 2.11 2.23 2.42 1.24 0.76 0.19 0.05 21 0.02 0.02 0.05 0.29 0.54 1.28 1.24 1.41 0.80 0.53 0.14 0.04 22 0.02 0.02 0.05 0.20 0.41 0.88 0.82 0.94 0.60 0.43 0.12 0.04 23 0.02 0.02 0.04 0.16 0.33 0.67 0.64 0.70 0.48 0.33 0.11 0.04

37

Figure 40 Mean isoprene ppbC from Danciger, 2005 – 2015, by month and hour

38

Figure 41 An example of isoprene modeling output from June 2012

Task 3: If the site selected is a not a current TCEQ monitoring site, the UT will prepare a new site to accommodate the trailer and monitor. In the event of either a new site or additions to an existing site, fencing and other necessary measures will secure the site, and if necessary, utilities will be connected to the site. UT will install, calibrate, and operate the Peak Performer 1 Reducing Compound Photometer (RCP) isoprene instrument, photosynthetically active radiation (PAR) instrument, and standard meteorological instrumentation (wind direction, wind speed, temperature) for a period of one calendar year. All instruments will collect data at an hourly or higher frequency. UT will be responsible for operating and maintaining the entire site according to the current TCEQ QAPP for Ambient Air Monitoring in Texas for SLAMS/Border, PM2.5, and PAMS Programs and the Photochemical Assessment Monitoring Stations (PAMS) Network QAPP for Ambient Air Monitoring in Texas. When an issue is noted by UT or TCEQ , including issues with the instruments, data, or monitoring trailer, UT or UT’s contractor will be on site no later than 24 hours after the concern (for example, a broken instrument or extraordinarily elevated sampled data) is realized to assess the issue. The UT will work with instrument manufacturers to troubleshoot and repair instrument malfunctions, which may include returning the equipment to the factory or installing a replacement part. All site preparation and instrument installation/ calibration/ operation activities will be summarized in a monthly monitoring operations report. The previous month's preliminary collected data and monitoring operations report will be delivered to the TCEQ Project Manager in an agreed upon electronic format by the 15th of every month or the following workday if on a weekend or holiday, starting no later than October 15, 2015. The monitoring will occur April 15, 2016 to October 15, 2016 and April 15, 2017 to October 15, 2017.

39

Task 4: UT will validate the collected isoprene and meteorological data according to the TCEQ Monitoring Division Standard Operating Procedure (SOP) DQRP-016 for meteorology with appropriate modifications for vertical wind measurements and with isoprene data. Additional instructions will be created using the Peak Performer 1 RCP Isoprene Analyzer Operations Manual that is specific to this project. Accuracy and precision metrics for all collected parameters will be calculated. The collected data will be flagged according to the quality-assured level of the data; including times of calibration, in-operation, and/or lost data. The validated data will be delivered to the TCEQ Project Manager in an agreed upon electronic format within 60 days of end of monitoring. In validating, analyzing, and summarizing data, UT will use SAS3 software 9.4, MS Excel, NOAA HYSPLIT, Golden Software Surfer 13, Google Earth Pro, or other software application approved by TCEQ, provided UT and TCEQ hold existing licensing rights to such software application. Task 5: UT will write a Final Report which will provide a comprehensive overview of activities undertaken and data collected and analyzed during the contract period. It will highlight major activities and key findings, provide pertinent analysis, describe encountered problems and associated corrective actions, and detail relevant statistics including parameter accuracy and precision. The report will include the data sources and raw data (provided on CD, flash memory drive, or FTP site). A draft of the final report will be submitted by January 25, 2018. The final report will be submitted by February 8, 2018.

5. Data Analyses

5.1 Ambient Isoprene and Other Data Figure 42 shows a time series graph of the one-hour time scale isoprene measurements taken at The Big Thicket site from March 2016 through October 2017, including the planned interruption over the 2016-2017 winter, outages owing to bad weather, and period of operation at other locations. The site recorded a maximum short term (5-minute) isoprene concentration on August 11, 2016 at 17:35 of 243 ppbC with a corresponding one-hour average of 227.6 ppbC for the hour starting at 17:00 CST. Figure 43 shows time series graphs for one-hour averaged isoprene, pyranometer (solar radiation), photosynthetic active radiation (PAR), humidity, and temperature, March 2016 to October 2016 and March 2017 to October 2017. Figure 43 is only intended to illustrate the time range of measurements and the range of values for each parameter. The diurnal patterns for isoprene concentrations by month, with afternoon peak concentrations, and nighttime near-zero concentrations appears in Figures 44, 45, and 46. Figure 44 shows the months from March through October 2016. In 2017, insufficient data were taken at Big Thicket in July and August, so only six months are available. Figure 45 shows the 2016 months excluding July and August, with the y-axis rescaled. Figure 46 shows the six months with available data in 2017. From Figures 45 and 46 can be drawn a conclusion that mean concentrations were generally lower in 2017 compared to 2016. A similar finding was made in 3 SAS is not an abbreviation. SAS Institute has developed an extensive suite of software for statistical analyses for which TCEQ and UT hold separate and individual existing licensing rights.

40

comparing the same six months in 2016 and 2017 at the TCEQ’s Nederland High School CAMS 1035 auto-GC, although no significant difference was observed with TCEQ’s Downtown Beaumont, which had similar mean concentrations between the two years, but much lower average concentrations than the other two sites.

41

Figure 42 Time series of Big Thicket 1-hour averaged isoprene initially coded “okay” and “ambient”, March 2016 – October 2016 and – March 2017 - October 2017

Co-location experiment

Lightning strike damage

42

Figure 43 Time series of Big Thicket 1-hour averaged isoprene, photosynthetic active radiation (PAR), pyranometer (total solar radiation), humidity, and temperature, March 2016 – October 2016 and – March 2017 - October 2017

43

Figure 44 Diurnal patterns for isoprene at Big Thicket in 2016 by month

Figure 45 Diurnal patterns for isoprene at Big Thicket in 2016 by month excluding July & August

44

Figure 46 Diurnal patterns for isoprene at Big Thicket in 2017 by month

5.2 Colocation experiment The Peak isoprene instrument and support gases were installed in late July 2017 at the TCEQ’s CAMS 1035 Nederland High School in Beaumont in order to compare measurements from the Peak instrument to isoprene measurements among the suite of 46 hydrocarbon species measured by the automated gas chromatograph (auto-GC) at the Nederland High station. The equipment ran at CAMS 1035 from July 31 to August 14, 2017 and was disconnected and moved to CAMS 1016 Lake Jackson (TCEQ) in Brazoria County for a similar Peak to auto-GC comparison. At Lake Jackson the instrument ran from August 16 to August 23, 2017. Figures 47 and 48 show the aerial images of the two colocation experiment sites with a circle of radius 1.0 mile drawn around each site. Photos to the Nederland High School site are at https://tinyurl.com/yc355gt9 and photos of the Lake Jackson site are at https://tinyurl.com/y8ggamcr (both accessed January 2018). On Thursday August 24, 2017, in advance of the approaching Hurricane Harvey, the Lake Jackson site was shut down. After the weather improved and conditions were safe, the Peak instrument was removed and reinstalled at the Big Thicket site on September 12, 2017. Figure 49 shows the comparison hourly time series for the Peak instrument and the coincident Nederland auto-GC. Because the auto-GC takes a 40-minute sample, only the first 8 five-minute Peak measurements were used each hour to calculate the mean for that hour. Figure 50 shows the Nederland and Peak isoprene comparison on August 7, 2017, in which the two match well in terms of the increases and decreases in concentration, but the Peak concentrations were all greater than the auto-GC concentrations. Figure 51 shows the linear regression for the Nederland isoprene against the Peak isoprene 40 min averages, in which the correlation is excellent but the Peak instrument concentrations were systematically higher than the coincident auto-GC concentrations. Figure 52 shows the comparison hourly time series for the Peak instrument and the coincident Lake Jackson auto-GC, and Figure 53 shows the linear regression comparison. Again, the correlation is excellent but the Peak instrument concentrations were systematically higher than the coincident auto-GC concentrations, although the comparison was better than at Nederland with the Lake Jackson slope closer to 1.0. The comparison at Lake Jackson was carried out for only 8 days, owing

45

to the imminent threat posed by Hurricane Harvey. The Lake Jackson site was shut down and secured on August 24. UT asked Orsat to comment on the observed differences between the Peak instrument and the TCEQ auto-GCs. Their response follows:

The differences between the Peak and AutoGC absolute concentrations can possibly be attributed to several factors. The data I have on AutoGC recoveries for isoprene are all at concentrations of approximately 30 ppbC and at this level the recoveries are 90-95%. However, we have seen that losses in these systems due to adsorption result in reduced recoveries as the concentration goes down. That is, if the system loses 5% at 30 ppbC to adsorption on steel and sorbents then it is possible that the percent recovery will go down as the loss remains constant but becomes a larger percentage of the expected as the concentration goes down. The Peak instrument also does not use any type of sample drier as the AutoGC does which could also account for some loses at low levels. In addition, no attempt was made to correlate possible differences in standards between the two systems although we did analyze both the isoprene standard from this year and last year against a propane standard we use in the lab to assure that they were within the certification spec supplied by the manufacturer. The blend tolerance for these standards is +/- 5%.

46

Figure 47 Nederland High School (TCEQ) in Jefferson County

Figure 48 Lake Jackson (TCEQ) in Brazoria County

47

Figure 49 Nederland High School isoprene vs Peak isoprene 40 min averages time series

Figure 50 Nederland and Peak isoprene comparison on August 7, 2017

48

Figure 51 Nederland High School isoprene vs Peak isoprene 40 min averages scatterplot/regression

Figure 52 Lake Jackson isoprene vs Peak isoprene 40 min averages time series

49

Figure 53 Lake Jackson isoprene vs Peak isoprene 40 min averages scatterplot/regression

5.3 Meteorological Data Collection Figure 54 compares 2016 and 2017 rainfall at the Big Thicket as measured by the National Weather Service WRRT2 site located 44 meters from the UT trailer. From early March through July, more rain fell in 2016 as opposed to 2017, but for most of the period from July through October, 2017 was wetter. In particular, Figure 54 shows the large increase in rainfall in 2017 from Hurricane Harvey in late-August 2017. Figure 54 Rainfall to date each year, 2016 and 2017 at Big Thicket

50

6. Conclusions and Recommendations Overall finding from this study to date are as follows:

• Higher isoprene concentrations have been measured at the Big Thicket site than at other sites in the State of Texas at auto-GC stations.

• Concentrations peak in the late afternoon, as predicted by models and shown by other monitors.

• Concentrations are higher under bright sunlight, as predicted by models and shown by other monitors.

• In colocation comparisons with two TCEQ monitoring stations, the agreement between the project Peak instrument and the TCEQ auto-GC were highly correlated, with higher concentrations recorded by the Peak instrument, which may be explained by its higher recovery for isoprene, which is a highly reactive gas.

If the TCEQ would identify specific questions to be answered though more data analysis, UT CEER would be happy to study them.

51

Appendix 1 Isoprene Analysis in Ambient Air SOP (by Orsat LLC)

A1.1 Description: The Peak PP1 Analyzer monitors ambient air for isoprene. The PP1 is a chromatographic analyzer that uses dual Unibeads 1S columns in an isothermal oven connected to an RCP (reducing compound photometer) detector with Nitrogen carrier gas. The PP1 injects a 1cc sample onto the columns every 5 minutes. The columns are held at 115°C and sample is heart- cut to the detector which is at 265°C. The remaining sample is back flushed off the column to vent for 100 seconds to prepare the columns for the next injection. The system will run a daily calibration and blank to ensure the system is operating within specification. These quality controls are run between 1:00 and 2:00 am and controlled by the Sutron data logger. The analytical result from each cycle is sent via RS232 to the Sutron data logger along with wind speed, wind direction, temperature, humidity, solar radiation, and pyrometer readings from the meteorological equipment. The site is equipped with a broadband cellular modem to allow the automatic polling of the data logger. The analyzer is additionally connected to a computer which allows remote access and control of the analyzer and collects the chromatographic data. The sample is collected from 3 meters above the ground using 1/8” SS Sulfinert® tubing attached to a 12” glass filter filled with glass wool for particulate filtration. The Sample line flow is controlled via a sample pump connected to the output of the sample manifold at 100mL/min. A sample at 50 mL/min is pulled from the manifold through the analyzer sample port and the residual 50mL/min flows by to vent. The dilution system is connected to the manifold with ~1ft 1/8” SS Sulfinert tubing and provides excess flow for calibration checks and blanks. Flow through the analyzer is controlled to 50mL using an MFC device. Excess flow on the manifold is controlled using a 1/8” adjustable rotameter. Cautions: The PP1 uses a mercury bed which releases small amounts of mercury to vent. The mercury scrubber should be changed out every 6 months. The RCP detector uses an UV light source and caution should be exercised when handling/viewing source. Flow should be maintained on the detector whenever heat is applied to the mercury bed to prevent significant damage to the detector. Safety: The PP1 has two headed zones and caution should be exercised when handling the oven and detector. The system utilizes several gas bottles and normal bottle handling practices should be used.

A1.2 Equipment:

• Peak Laboratories PP1 Analyzer Model 910-32 • Merlin MicroScience Dilution system Model MMSD-VOC • Computer • Sutron Xpert 2 Data logger • Sampling pump • SS sample line

52

• Standard (1 ppmV) in summa canister Consumables: • Nitrogen gas (PP1 carrier) • Air (valve actuator) • N2 or Air (dilution system) • 1ppm Isoprene standard

A1.3 System Settings:

• Support Gases: All bottle gas output can be set between 70-80 psig. Nitrogen carrier bottles can be ganged together so bottles can be changed out without disrupting system operation.

• PP1 settings: The PP1 oven temperature should be set to 115°C while the detector is held constant at 265°C. The PP1 lamp signal should be between 1200 and 2200mV. Valve timing on PP1 is set to inject sample at 1 second, the detector signal is zeroed at 50 seconds, and the injection valve returned to back-flush at 80 seconds. The PP1 run ends at 300 seconds and immediately begins the next run. All results were reported as ppbC.

A1.4 Quality Control: Calibration curves are run at the initial start of sampling and at the end. Additionally calibrations are performed if any significant maintenance is performed. Quality Control samples are used to ensure that the data produced from the analyzer are of a known quality, and consistent throughout the analytical process. Nightly quality control checks along with analytical blanks are run to insure that the response of the instrument remains consistent and does not vary. Quality control limits are listed in Table 4. Table 4 Quality Controls

Quality Control Sample Limits Nightly Check Standard ±25% Analytical Blank < 2 ppbC Calibration Curve < 10% RSD

A1.5 Operator Responsibilities: The site operator is responsible for monitoring and inspecting all parameters in order to maintain system operation and data collection. Routine site checks: • Daily remote checks- Operator will review calibrations daily. Spot checking the

chromatograms to ensure no interferences and proper integration is recommended. File sorting should be done as needed to keep data storage areas free of clutter.

• Weekly onsite checks- Operator will visit the site once a week and check gas bottles for

pressure. Standard canister pressure and dilution system pressures should be checked to

53

verify calibration levels. Pad site should be checked for plant growth and safety issues. Additional site visits may be required depending on the results of routine site checks and quality control sample results. The site operator is required to investigate and correct problems which result in a failure of quality control data to meet the quality objectives and correct problems with communications when necessary.

A1.6 System Maintenance: Operator will change out UV light source once signal has dropped below 1200mv. Mercury bed and mercury scrubber should be changed out every 24 months. Standard canister should be refilled before canister pressure drops below 16 psig to ensure dilutor has proper supply pressure. Gas bottles should be changed out before bottle reaches 250 psig.

A1.7 System Calibration: The dilution system will supply a daily calibration check level of nominally 50 ppb used to check the 3 point calibration response factor. The 3 point calibration is run at 13 ppbC, 99 ppbC, and 168 ppbC and is run anytime major changes to the instrument are made as well as at startup and shutdown. Figure 55 shows the dilution system set-points used for this calibration curve and the resulting calculation of response factor used in the PP1 analytical system. Figure 56 shows the chromatograms for the three point curve. Figure 57 shows the nightly check standard, blank and a typical ambient chromatogram. Appendix 6.A includes the certificate of analysis for the isoprene stock standard used for this testing.

A1.8 References: Peak PP1 RCP manual

54

Figure 55 Dilution System Calibration Curve

Figure 56 Calibration Curve Chromatograms - High, Mid and Low points

.

Figure 57 Nightly check standard (48 ppbC), blank (0.8 ppbC) and typical ambient sample (32.5 ppbC)

55

A1.9 Airgas certification

56

Appendix 2 Tree Survey

Tree counts in 28 sampled 0.1 acre locations UT has examined the tree survey conducted by Parker Forestry Consultants (PFC). The PFC report is attached. Based on a sampling of 28 tracts of land, each 0.1 acre in size, UT offers some extrapolations to estimate a tree census within one half mile of the former UT isoprene monitoring station. The front cover of the PFC report shows a pie chart of the 19 different tree species identified and counted from the sampling, and this graph is repeated in Figure 58. Figure 58 Summary of tree species sampled near the Big Thicket monitoring station

Although 19 tree species were identified, six species – Black Gum, Holly, Magnolia, Oak, Pine, and Sweet Gum – represented 91 percent of the trees counted. The average count per lot for these and the total tree count average per lot is shown in Table 1, along with other statistics.

57

Table 5 Tree count statistics by type, 0.1 acres, for six major species

Species BlackGum Holly Magnolia Oak Pine SweetGum Grand Total Mean count 2.96 3.32 3.21 14.07 7.50 4.96 39.79

Standard error 0.11 0.16 0.10 0.34 0.35 0.25 0.65 Standard deviation 3.08 4.28 2.59 9.20 9.37 6.83 17.62

Maximum 10 17 10 33 41 29 96 Minimum 0 0 0 1 0 0 16

Figures 59, 60, and 61 show histograms of the number of the trees in the 28 tracts for the three most common species: Oak, Pine, and Sweet Gum. Figure 62 shows the histogram for total trees per lot. Figure 59 Oak histogram – how many lots (y-axis) had this many trees (x-axis)

58

Figure 60 Pine histogram– how many lots (y-axis) had this many trees (x-axis)

Figure 61 Sweet Gum histogram– how many lots (y-axis) had this many trees (x-axis)

59

Figure 62 Total trees count histogram– how many lots (y-axis) had this many trees (x-axis)

Extrapolation The six tree species shown in Table 1 accounted for 91 percent of all trees counted. The survey locations were 0.1 acre, or about 4,400 square feet, which is a little smaller than a single family lot. More specifically, an acre is 43,560 square feet, and there are 640 acres in a square mile, and 247.1 acres in a square kilometer. Overall, an average of 40 trees per 0.1 acre were counted around the monitoring site out to a 0.5 miles radius. Extrapolating this using no other information suggests 400 trees per acre, or 256,000 trees per square mile, or 99,000 trees per square kilometer. For oaks, 14 trees per 0.1 acre, with no other information, extrapolates to 140 per acre, 90,000 per square mile, or 34,600 per square kilometer. For sweet gum, 5 trees per 0.1 acre, with no other information, extrapolates to 50 per acre, 32,000 per square mile, or 12,350 per square kilometer. Additional information to refine these extrapolations could come from looking at how the tree counted were distributed across the area. For example, if there were a clear pattern for more trees of a certain species in one part of the area of interest that had relatively few sampling lots in it, we might suspect that in averaging over all sampling lots equally weighted would underestimate the true tree count for that species. If a systematic pattern for the distribution of a tree species were to be found, then one means to address it would be divide the sampling region up into sub-regions and average with sub-regions and then average across all sub-region, weighting each by its size. A more rigorous alternative would be to perform a contouring across the area, using, for example, Gaussian kriging, and then integrating across the region. This integration is easily accomplished by averaging over the array of grid points created in kriging. Both of these techniques were tried with the total tree data, but no significant differences were produced, as explained below. UT closely examined the distribution of tree count and two tree species over the 0.5 mile radius, 0.785 square mile, 2.0 square kilometer area surveyed by PFC. Figure 63 shows a linear

60

regression fit to the difference in total tree count between pairs of lots versus the distance between pairs of lots. For the 28*(28–1)/2 = 378 pairs of lots, there was no clear relationship based on distance, as shown by the flat line fit to the graph in Figure 63 with insignificant slope, suggesting a near random distribution of tree counts. Similarly, Figure 64 shows the shows the insignificant linear regression fit to the difference in oak tree count between pairs of lots versus the distance between pairs of lots and Figure 65 shows the same for sweet gum trees. The relationship appears better for sweet gum than for oak, but the correlation is still very low at 0.1. Thus, it seems that the simple average across the survey area is sufficient to represent the distribution for total tree count, oaks, and sweet gum. A similar analysis can be done for other species upon request.

61

Figure 63 Linear regression for difference in total tree count between pairs of lots versus the distance between pairs of lots

62

Figure 64 Linear regression for difference in oak tree count between pairs of lots versus the distance between pairs of lots

63

Figure 65 Linear regression for difference in sweet gum tree count between pairs of lots versus the distance between pairs of lots

64

Parker Forestry University of Texas Isoprene Study Submitted Report: Big Thicket National Preserve January 2018

65

Parker Forestry Consultants 411 N. Pine

Woodville, TX 75979 409-283-5413 telephone

409-283-5414 fax

pfc@parkerforestry.com Isoprene Study

Big Thicket National

Preserve Tyler County, TX

Field work commenced on December 28, 2017 and concluded on January 8, 2018.

Project: The University of Texas Center for Energy and Environments Resources contracted with Parker Forestry Consultants through Dr. David W. Sullivan to identify and tally all trees (greater than fifteen feet tall) within twenty-eight randomly placed one-tenth acre plots in the Big Thicket National Preserve

Purpose: The purpose of the study is to monitor isoprene emission by native vegetation in a natural forest.

Location: All plots were randomly generated using a ARCMAP GIS software tool called Create Random Points. The plots were randomly placed within a one-half mile radius of a monitoring station that is operated by the University of Texas. The location of the monitoring station is Latitude 30.544632°, Longitude –94.346562°. The plots were located within the boundaries of the Big Thicket National Preserve. Foresters were able to drive to the monitoring station with a vehicle and navigate to the plots using GPS equipment loaded with the predetermined points. Foresters were also provided aerial field maps. When the correct plot location was reached by foot, a plot center was established with flagging pushed into the ground with the plot number, date and foresters initials.

Method Determination: The project was completed in two stages:

1. Initially two one-tenth acre plots were tallied and Parker Forestry sent a short report and the information gathered to Dr. Dave Sullivan at The University of Texas at Austin. Dr. Sullivan provided feedback on the format and level of detail he needed.

2. After consulting on the first stage with Dr. Sullivan, the second stage was completed. In all, twenty-eight one-tenth acre plots were tallied.

66

Field Method: All trees over fifteen feet in total height and .6 inches and over in diameter breast height (dbh) were tallied by foresters working in pairs. All qualifying tree species within the one-tenth acre circular plots (radius of 37.2 feet) were measured. The foresters tallied the trees using Husky field computers with a simple method file to tally tree by species, diameter in inches and total height in feet. Total height was measured from ground level to the tip of the tallest branch. Foresters used Haglof DMEs, to determine the radius of the circular plots. A Haglof DME uses ultrasonic pulses to determine distance, allowing the forester to walk 360° around the transponder and receive accurate measurements at the push of a button-- http://www.haglofsweden.com/index.php/en/products/instruments/height/328-dme

One forester in each group stood at plot center and tallied on the hand-held computer; the forester in the plot center helped determine tree heights and species while the other forester measured tree diameter with either a caliper or diameter tape (marked in 1/10th inch increments), depending upon the tree size. Trees over eighteen inches in diameter were measure using a forester’s diameter tape. Total height was measured with a sixteen foot collapsing pole (for short stems), a hand held clintometer or a TruPulse 360B laser clintometer. The TruPulse is a professional laser rangefinder which measures distances and heights and offers in-scope data display with power optics-- http://www.lasertech.com/TruPulse-Laser-Rangefinder.aspx. If needed, each forester carried a logger’s tape to determine plot distance and distance from larger trees if the clintometer was used for total heights.

Foresters were able to determine tree species by bark and bud as well as leaves on the ground. If in doubt, the twig was brought back to the offer for other’s opinion. Underbrush species, such as Yaupon, Chinese Privet and Titi were not tallied but the number of stems within the one-tenth acre plot was estimated. Foresters also described the plots location’s general timber type.

Foresters: Tim A. Parker, Jeffrey D. Parker, Colton McKee, and Carl Parker completed the field work. Keelin Parker negotiated the contract, provided job specifications to the foresters and computer the final tally and wrote the report. All members of the crew (except Carl Parker) and office staff have a degree in forestry, GIS, and/or wildlife. Carl Parker is college senior studying wildlife biology.

Summary:

• 28 one-tenth acre circular plots were randomly placed within one-half mile radius of the UT Monitoring station;

• Trees >15’ tall were tallied by species, diameter in inches at diameter breast height (4.5” above ground level), and total height in feet;

• Foresters qualitatively characterized the undergrowth by species, number of stems, and height;

• Foresters described the forest type surrounding the plot; • All measurements were made by professional foresters using electronic and hand-

held forestry equipment. • Foresters have worked in Tyler County for many years and have knowledge of the

local species as well as bark and bud characteristics.

67

Tally of All Trees Over 15' in Total Height 28- 1/10th Acre Circular Plots Randomly Placed within .5 Miles of Monitoring Station Latitude 30.544632/Longitude -94.346562 Big Thicket National Preserve Tyler County, TX December & January 2018 Tallied by Parker Forestry Consultants

Table 6 University of Texas Isoprene Study ‐ January 2018

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

1 1 Sweetgum 1 8 55 1 2 Holly 1 3 32 1 3 Holly 1 4 36 1 4 Holly 1 4 25 1 5 Holly 1 3 29 1 6 Holly 1 7 26 1 7 Oak 1 10 47 1 8 BlackGum 1 6 44 1 9 BlackGum 1 6 31 1 10 Holly 1 6 34 1 11 Holly 1 4 36 1 12 Holly 1 6 33 1 13 Holly 1 4 23 1 14 BlackGum 1 5 29 1 15 Holly 1 5 26 1 16 Holly 1 6 24 1 17 Oak 1 17 49 1 18 BlackGum 1 9 43 1 19 Holly 1 3 25 2 1 BlackGum 1 6 34 2 2 Pine 1 2 21 2 3 Magnolia 1 5 27 2 4 Sweetgum 1 5 31 2 5 Pine 1 2 18 2 6 Sweetgum 1 7 43 2 7 Sweetgum 1 3 24 2 8 BlackGum 1 4 33 2 9 BlackGum 1 7 32 2 10 BlackGum 1 7 32 2 11 Sweetgum 1 4 34 2 12 BlackGum 1 2 22 2 13 BlackGum 1 1 16 2 14 Sweetgum 1 2 16 2 15 Holly 1 5 26 2 16 Oak 1 2 24

University of Texas Isoprene Study ‐ January 2018

68

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

2 17 Magnolia 1 6 28 2 18 Magnolia 1 3 22 2 19 Magnolia 1 2 18 2 20 Sweetgum 1 6 35 2 21 Sweetgum 1 4 26 2 22 Oak 1 5 21 2 23 Sweetgum 1 7 41 2 24 Oak 1 1 17 2 25 Sumac 1 1 18 2 26 Sweetgum 1 6 27 2 27 Sumac 1 2 18 2 28 Sumac 1 2 18 2 29 Sweetgum 1 2 16 2 30 Oak 1 2 20 2 31 Oak 1 1 17 2 32 Oak 1 2 20 2 33 Oak 1 3 21 2 34 Pine 1 7 34 2 35 Pine 1 11 52 2 36 Pine 1 4 32 2 37 Sweetgum 1 5 31 2 38 Pine 1 11 44 2 39 Sweetgum 1 5 30 2 40 Oak 1 5 34 2 41 Sweetgum 1 3 26 2 42 BlackGum 1 3 23 2 43 Oak 1 1 17 2 44 Oak 1 2 17 2 45 Oak 1 1 18 2 46 Oak 1 2 21 2 47 BlackGum 1 4 24 3 1 Oak 1 9 54 3 2 Oak 1 8 39 3 3 BlackGum 1 3 40 3 4 Oak 1 9 51 3 5 Oak 1 5 32 3 6 Magnolia 1 6 38 3 7 Magnolia 1 5 39 3 8 Magnolia 1 7 39 3 9 Magnolia 1 9 45 3 10 Oak 1 1 17 3 11 Holly 1 5 32 3 12 Holly 1 5 34 3 13 Holly 1 5 33

University of Texas Isoprene Study ‐ January 2018

69

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

3 14 Oak 1 2 19 3 15 Redbay 1 3 25 3 16 Holly 1 7 33 3 17 BlackGum 1 3 19 3 18 Sweetgum 1 4 27 3 19 Oak 1 9 47 3 20 Maple 1 3 28 3 21 Oak 1 7 39 3 22 BlackGum 1 7 42 3 23 Oak 1 11 61 4 1 Pine 1 12 49 4 2 Pine 1 9 44 4 3 Pine 1 5 36 4 4 Oak 1 5 19 4 5 BlackGum 1 8 40 4 6 Oak 1 7 35 4 7 Oak 1 2 29 4 8 Oak 1 2 28 4 9 Oak 1 5 26 4 10 Sweetgum 1 2 20 4 11 Oak 1 6 34 4 12 Oak 1 3 32 4 13 BlackGum 1 10 70 4 14 BlackGum 1 7 30 4 15 Oak 1 5 34 4 16 Oak 1 4 32 4 17 Oak 1 2 24 4 18 Oak 1 3 20 4 19 Oak 1 2 19 4 20 Magnolia 1 2 24 4 21 Oak 1 6 28 4 22 Oak 1 10 37 4 23 Oak 1 11 41 5 1 Sweetgum 1 8 53 5 2 Oak 1 3 17 5 3 Pine 1 1 17 5 4 Maple 1 2 23 5 5 Maple 1 2 25 5 6 Holly 1 2 19 5 7 Sweetgum 1 2 21 5 8 Oak 1 10 52 5 9 Holly 1 5 32 5 10 Maple 1 2 24 5 11 Oak 1 2 20

University of Texas Isoprene Study ‐ January 2018

70

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

5 12 Sweetgum 1 8 52 5 13 Pine 1 4 29 5 14 Sweetgum 1 8 54 5 15 Pine 1 3 23 5 16 Pine 1 3 23 5 17 Sweetgum 1 9 54 5 18 Pine 1 2 21 5 19 Sweetgum 1 4 25 5 20 Sweetgum 1 4 26 5 21 Oak 1 12 55 5 22 Oak 1 2 19 5 23 Sweetgum 1 8 52 5 24 Oak 1 2 20 5 25 Oak 1 8 42 5 26 Oak 1 9 43 5 27 Oak 1 11 56 5 28 Pine 1 2 21 5 29 Oak 1 6 39 5 30 Oak 1 8 42 5 31 Pine 1 5 37 5 32 Oak 1 6 36 5 33 Maple 1 5 35 5 34 Maple 1 4 39 5 35 Pine 1 4 35 5 36 Pine 1 9 39 5 37 Maple 1 4 24 5 38 Maple 1 2 36 5 39 Maple 1 1 23 5 40 Oak 1 8 32 5 41 Oak 1 3 32 6 1 Magnolia 1 10 37 6 2 Magnolia 1 4 32 6 3 Sweetleaf 1 2 21 6 4 BlackGum 1 4 36 6 5 BlackGum 1 4 19 6 6 BlackGum 1 10 55 6 7 Sweetleaf 1 2 19 6 8 BlackGum 1 7 49 6 9 BlackGum 1 4 26 6 10 BlackGum 1 9 59 6 11 BlackGum 1 5 40 6 12 BlackGum 1 6 37 6 13 Sweetgum 1 6 39 6 14 Holly 1 3 22

University of Texas Isoprene Study ‐ January 2018

71

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

6 15 Sweetgum 1 9 52 6 16 Magnolia 1 11 50 6 17 Magnolia 1 6 39 6 18 BlackGum 1 12 56 6 19 Holly 1 3 21 6 20 BlackGum 1 11 61 6 21 Magnolia 1 6 36 6 22 Magnolia 1 8 47 6 23 Oak 1 18 72 6 24 Pine 1 18 90 6 25 Sweetgum 1 3 33 6 26 Pine 1 7 45 6 27 Pine 1 23 88 7 1 Oak 1 2 15 7 2 Oak 1 2 19 7 3 Oak 1 6 30 7 4 Pine 1 2 19 7 5 Pine 1 3 20 7 6 Pine 1 7 45 7 7 Oak 1 4 33 7 8 Oak 1 4 26 7 9 Pine 1 3 28 7 10 BlackGum 1 17 52 7 11 Magnolia 1 8 38 7 12 Pine 1 5 40 7 13 Magnolia 1 1 17 7 14 Oak 1 7 33 7 15 Magnolia 1 8 35 7 16 Magnolia 1 2 24 7 17 Holly 1 2 18 7 18 Pine 1 7 44 7 19 Magnolia 1 6 44 7 20 Magnolia 1 3 32 7 21 Magnolia 1 1 15 7 22 Pine 1 9 45 7 23 Pine 1 8 45 7 24 Pine 1 8 46 7 25 Pine 1 3 35 7 26 Pine 1 6 40 7 27 Pine 1 4 40 7 28 Pine 1 2 21 7 29 Pine 1 6 44 7 30 Pine 1 5 40 7 31 Pine 1 6 44

University of Texas Isoprene Study ‐ January 2018

72

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

7 32 Pine 1 4 35 7 33 Magnolia 1 2 28 7 34 Magnolia 1 2 27 7 35 Magnolia 1 2 20 7 36 Pine 1 5 36 7 37 BlackGum 1 3 24 7 38 Sweetgum 1 6 34 7 39 Pine 1 1 16 7 40 Pine 1 3 28 7 41 Pine 1 8 45 7 42 Redbay 1 2 19 7 43 Pine 1 6 28 7 44 Oak 1 6 33 7 45 Pine 1 2 30 7 46 Pine 1 4 33 7 47 Oak 1 1 24 8 1 Pine 1 20 97 8 2 Oak 1 8 47 8 3 BlackGum 1 11 56 8 4 Oak 1 12 55 8 5 Oak 1 8 48 8 6 BlackGum 1 2 18 8 7 Oak 1 9 45 8 8 Oak 1 10 43 8 9 Oak 1 11 49 8 10 Oak 1 9 45 8 11 Pine 1 10 76 8 12 Pine 1 20 97 8 13 Oak 1 6 35 8 14 BlackGum 1 4 32 8 15 Oak 1 6 31 8 16 Pine 1 15 99 9 1 Magnolia 1 6 37 9 2 Oak 1 1 17 9 3 Oak 1 2 22 9 4 Magnolia 1 7 32 9 5 Oak 1 2 32 9 6 Maple 1 8 46 9 7 Sweetgum 1 13 51 9 8 BlackGum 1 5 45 9 9 Sweetgum 1 3 34 9 10 Maple 1 5 39 9 11 Sweetgum 1 7 46 9 12 Oak 1 3 19

University of Texas Isoprene Study ‐ January 2018

73

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

9 13 Sweetgum 1 2 24 9 14 Oak 1 4 31 9 15 Oak 1 2 30 9 16 Sweetgum 1 11 63 9 17 Pine 1 19 85 9 18 Oak 1 2 26 9 19 Oak 1 3 43 9 20 Pine 1 8 62 9 21 Sweetgum 1 2 22 9 22 Sweetgum 1 2 23 9 23 Sweetgum 1 2 24 9 24 Magnolia 1 1 16 9 25 Pine 1 4 41 9 26 Pine 1 4 37 9 27 Oak 1 2 21 9 28 Oak 1 1 15 9 29 Pine 1 6 49 9 30 Pine 1 16 69 9 31 Oak 1 3 22 9 32 Oak 1 3 22 9 33 Oak 1 2 16 9 34 Sweetgum 1 12 55 9 35 Pine 1 13 66 9 36 Oak 1 2 22 9 37 Oak 1 6 37 9 38 Oak 1 5 45 9 39 Oak 1 2 16 9 40 Oak 1 7 42 9 41 Pine 1 10 71 9 42 Oak 1 2 16 9 43 Pine 1 14 72 9 44 Maple 1 3 34 9 45 Sweetgum 1 3 45 9 46 Magnolia 1 3 25 9 47 Oak 1 3 23 9 48 Pine 1 12 77 9 49 Oak 1 3 42 9 50 Oak 1 2 21 9 51 Magnolia 1 2 24 9 52 Magnolia 1 2 23 9 53 Oak 1 3 31 9 54 Pine 1 7 61 9 55 Pine 1 6 55 9 56 Pine 1 5 43

University of Texas Isoprene Study ‐ January 2018

74

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

9 57 Sweetgum 1 4 34 9 58 Sweetgum 1 7 46 9 59 Sweetgum 1 6 53 9 60 Pine 1 5 45 9 61 Sweetgum 1 2 16 9 62 Oak 1 1 16 9 63 Holly 1 10 55 9 64 Oak 1 11 65 9 65 Oak 1 2 19 9 66 Oak 1 4 27 9 67 Oak 1 3 17 9 68 Oak 1 5 25 9 69 Oak 1 4 27 9 70 Pine 1 7 44 9 71 Oak 1 3 20

10 1 Oak 1 3 19 10 2 Oak 1 6 19 10 3 Oak 1 4 21 10 4 Oak 1 1 15 10 5 Oak 1 5 30 10 6 Oak 1 3 21 10 7 Oak 1 1 16 10 8 Oak 1 1 15 10 9 Oak 1 3 30 10 10 Oak 1 4 28 10 11 Oak 1 6 32 10 12 Oak 1 5 26 10 13 Sweetgum 1 2 20 10 14 Oak 1 2 32 10 15 Oak 1 2 20 10 16 Oak 1 2 19 10 17 Oak 1 1 19 10 18 Oak 1 7 24 10 19 Oak 1 1 18 10 20 Pine 1 4 35 10 21 Oak 1 13 58 10 22 Oak 1 13 57 10 23 Oak 1 4 38 10 24 Oak 1 12 58 10 25 BlackGum 1 2 17 10 26 Oak 1 10 47 10 27 Oak 1 2 22 10 28 Oak 1 14 58 10 29 Sweetleaf 1 1 17

University of Texas Isoprene Study ‐ January 2018

75

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

10 30 Oak 1 4 24 10 31 Oak 1 1 23 10 32 Pine 1 1 16 10 33 Magnolia 1 4 27 10 34 Magnolia 1 5 29 10 35 Oak 1 10 37 10 36 Oak 1 1 15 10 37 Oak 1 3 25 11 1 Dogwood 1 6 33 11 2 Pine 1 1 17 11 3 Holly 1 2 16 11 4 Sweetgum 1 3 20 11 5 Magnolia 1 7 40 11 6 Magnolia 1 2 22 11 7 Magnolia 1 3 18 11 8 Oak 1 11 45 11 9 Oak 1 12 51 11 10 Maple 1 3 21 11 11 Maple 1 6 45 11 12 Maple 1 2 21 11 13 Maple 1 3 27 11 14 Maple 1 1 15 11 15 Pine 1 19 90 11 16 Magnolia 1 2 15 11 17 Oak 1 1 18 11 18 Oak 1 3 24 11 19 Pine 1 24 87 11 20 Oak 1 1 18 11 21 Oak 1 2 18 11 22 Oak 1 2 20 11 23 Sweetgum 1 4 25 11 24 Holly 1 4 26 11 25 Pine 1 5 30 11 26 Pine 1 11 76 11 27 Oak 1 6 29 11 28 Magnolia 1 1 18 11 29 Oak 1 4 35 11 30 Oak 1 1 17 11 31 Holly 1 2 17 11 32 Pine 1 6 36 11 33 Pine 1 2 20 11 34 Oak 1 2 21 11 35 Holly 1 2 18 11 36 Oak 1 9 50

University of Texas Isoprene Study ‐ January 2018

76

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

11 37 Pine 1 2 19 11 38 Oak 1 7 36 11 39 Holly 1 3 23 11 40 Pine 1 1 20 11 41 Pine 1 4 34 11 42 Pine 1 8 45 11 43 Pine 1 7 46 11 44 Oak 1 5 34 11 45 Sweetgum 1 5 27 11 46 Oak 1 2 15 11 47 Pine 1 4 25 11 48 Pine 1 13 74 11 49 Sweetgum 1 3 24 11 50 Pine 1 2 23 11 51 Redbay 1 7 34 12 1 Pine 1 6 42 12 2 Oak 1 3 41 12 3 Oak 1 2 29 12 4 Oak 1 4 40 12 5 Oak 1 4 41 12 6 Sparkleberry 1 2 20 12 7 Sparkleberry 1 1 25 12 8 BlackGum 1 5 44 12 9 Oak 1 1 17 12 10 Oak 1 1 19 12 11 Oak 1 8 38 12 12 Pine 1 7 52 12 13 Pine 1 11 66 12 14 Pine 1 12 70 12 15 Oak 1 2 31 12 16 Oak 1 3 27 12 17 Holly 1 3 32 12 18 Sweetleaf 1 2 26 12 19 Dogwood 1 3 19 12 20 Oak 1 2 29 12 21 Sweetgum 1 12 60 12 22 Oak 1 2 25 12 23 Oak 1 3 29 12 24 Oak 1 3 27 12 25 Oak 1 2 24 12 26 Oak 1 5 35 12 27 Pine 1 10 49 12 28 Oak 1 2 32 12 29 Oak 1 2 30

University of Texas Isoprene Study ‐ January 2018

77

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

12 30 Dogwood 1 3 18 12 31 Sweetleaf 1 2 22 12 32 Pine 1 7 44 12 33 Oak 1 8 41 12 34 Oak 1 2 31 12 35 Oak 1 3 31 12 36 Holly 1 2 17 12 37 Holly 1 4 31 12 38 Pine 1 4 36 12 39 Magnolia 1 7 39 12 40 Oak 1 3 32 12 41 Oak 1 1 25 12 42 Holly 1 7 42 12 43 Holly 1 5 38 12 44 Magnolia 1 11 50 12 45 Oak 1 1 16 12 46 BlackGum 1 10 45 13 1 Magnolia 1 11 58 13 2 BlackGum 1 6 34 13 3 Magnolia 1 7 41 13 4 BlackGum 1 4 25 13 5 BlackGum 1 5 37 13 6 BlackGum 1 6 41 13 7 BlackGum 1 10 57 13 8 Magnolia 1 8 45 13 9 Magnolia 1 6 36 13 10 Oak 1 17 61 13 11 BlackGum 1 4 37 13 12 Redbay 1 2 23 13 13 BlackGum 1 2 17 13 14 Sweetgum 1 8 37 13 15 BlackGum 1 10 27 13 16 BlackGum 1 20 49 13 17 Maple 1 3 36 13 18 Magnolia 1 12 49 13 19 Magnolia 1 4 31 13 20 Redbay 1 2 18 13 21 BlackGum 1 8 46 14 1 Oak 1 3 22 14 2 Maple 1 4 31 14 3 Holly 1 3 29 14 4 Dogwood 1 3 23 14 5 Hickory 1 14 75 14 6 BlackGum 1 4 32

University of Texas Isoprene Study ‐ January 2018

78

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

14 7 BlackGum 1 3 26 14 8 Ironwood 1 1 23 14 9 Holly 1 2 25 14 10 Holly 1 2 25 14 11 Holly 1 8 37 14 12 BlackGum 1 5 33 14 13 Ironwood 1 4 32 14 14 Sweetgum 1 12 85 14 15 Oak 1 17 73 14 16 Maple 1 4 37 14 17 Holly 1 3 26 14 18 Holly 1 3 20 14 19 Holly 1 4 35 14 20 Sweetgum 1 6 45 14 21 Ironwood 1 3 23 14 22 Holly 1 2 22 14 23 Ironwood 1 3 23 14 24 Oak 1 4 38 14 25 Oak 1 5 37 14 26 BlackGum 1 15 75 14 27 Ironwood 1 3 16 14 28 Maple 1 4 24 14 29 Ironwood 1 2 22 14 30 Oak 1 4 32 14 31 Sweetleaf 1 2 17 14 32 Holly 1 3 19 14 33 Oak 1 10 60 14 34 Hickory 1 8 62 14 35 Pine 1 21 90 14 36 Ironwood 1 2 34 14 37 Dogwood 1 3 29 14 38 Holly 1 3 24 14 39 Holly 1 1 21 14 40 BlackGum 1 5 33 15 1 BlackGum 1 2 34 15 2 BlackGum 1 5 16 15 3 Sweetgum 1 5 35 15 4 Sweetgum 1 18 80 15 5 Magnolia 1 5 31 15 6 Oak 1 1 17 15 7 Oak 1 2 21 15 8 Redbay 1 2 17 15 9 Sweetgum 1 9 57 15 10 Sweetgum 1 9 55

University of Texas Isoprene Study ‐ January 2018

79

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

15 11 Sweetgum 1 5 35 15 12 Oak 1 3 21 15 13 BlackGum 1 4 32 15 14 BlackGum 1 9 45 15 15 Redbay 1 2 17 15 16 Magnolia 1 10 43 15 17 Magnolia 1 2 27 15 18 Magnolia 1 4 21 15 19 BlackGum 1 7 51 15 20 Pine 1 4 27 15 21 BlackGum 1 7 52 15 22 BlackGum 1 5 25 16 1 Oak 1 14 65 16 2 Oak 1 1 17 16 3 Magnolia 1 5 40 16 4 Oak 1 7 47 16 5 Oak 1 8 45 16 6 Pine 1 23 89 16 7 Sweetgum 1 2 17 16 8 Sweetgum 1 3 17 16 9 Sweetgum 1 3 17 16 10 Sweetgum 1 2 16 16 11 Sweetgum 1 4 17 16 12 Sweetgum 1 3 23 16 13 Holly 1 8 43 16 14 Pine 1 14 87 16 15 Sweetgum 1 3 22 16 16 Pine 1 12 82 16 17 Sweetgum 1 2 26 16 18 Holly 1 4 25 16 19 Holly 1 7 37 16 20 Pine 1 16 93 16 21 Oak 1 9 52 16 22 Pine 1 15 95 16 23 Magnolia 1 2 24 16 24 Pine 1 5 45 16 25 Pine 1 14 96 16 26 Pine 1 10 87 16 27 Oak 1 5 21 16 28 Pine 1 13 76 16 29 Oak 1 2 23 16 30 Sweetgum 1 3 21 16 31 Pine 1 5 35 16 32 Pine 1 6 43

University of Texas Isoprene Study ‐ January 2018

80

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

16 33 Pine 1 6 43 16 34 Pine 1 7 47 16 35 Oak 1 3 17 16 36 Oak 1 7 45 16 37 Pine 1 7 53 16 38 Pine 1 4 31 16 39 Pine 1 2 25 16 40 Pine 1 3 26 16 41 Oak 1 1 26 16 42 Oak 1 3 25 16 43 Pine 1 2 17 16 44 Pine 1 3 21 16 45 Pine 1 4 41 16 46 Pine 1 1 16 16 47 Pine 1 3 27 16 48 Oak 1 3 17 17 1 Pine 1 9 54 17 2 Pine 1 9 56 17 3 Holly 1 5 37 17 4 Holly 1 4 27 17 5 Holly 1 3 24 17 6 Pine 1 3 17 17 7 Pine 1 2 21 17 8 Holly 1 3 21 17 9 Pine 1 4 22 17 10 Oak 1 3 23 17 11 Sweetgum 1 6 36 17 12 Holly 1 3 33 17 13 Holly 1 5 37 17 14 Holly 1 5 38 17 15 Pine 1 4 33 17 16 Oak 1 3 33 17 17 Sweetgum 1 4 26 17 18 Oak 1 10 47 17 19 Oak 1 3 23 17 20 Pine 1 2 19 17 21 Pine 1 2 19 17 22 Oak 1 10 34 17 23 Pine 1 4 39 17 24 Oak 1 3 36 17 25 Pine 1 5 35 17 26 Sweetgum 1 4 25 17 27 Pine 1 6 50 17 28 Pine 1 5 52

University of Texas Isoprene Study ‐ January 2018

81

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

17 29 Pine 1 3 24 17 30 Sweetgum 1 2 19 17 31 Sweetgum 1 4 34 17 32 Pine 1 12 67 17 33 Oak 1 5 35 17 34 Oak 1 3 25 17 35 Pine 1 5 35 17 36 Sweetgum 1 5 27 17 37 Oak 1 2 17 17 38 Pine 1 7 47 17 39 Maple 1 6 44 17 40 Oak 1 1 16 17 41 Pine 1 6 56 17 42 Magnolia 1 2 19 17 43 Oak 1 2 19 17 44 Pine 1 10 72 17 45 Oak 1 2 22 17 46 Oak 1 3 27 17 47 Pine 1 3 27 17 48 Pine 1 14 79 17 49 Oak 1 1 15 17 50 Pine 1 2 23 17 51 Oak 1 3 23 17 52 Pine 1 6 61 17 53 Sweetgum 1 3 21 17 54 Oak 1 7 35 17 55 Cherry Laurel 1 2 17 17 56 Oak 1 1 17 17 57 Oak 1 9 53 17 58 Pine 1 2 16 17 59 Pine 1 2 16 17 60 Pine 1 2 17 17 61 Pine 1 3 24 17 62 Sweetgum 1 2 19 17 63 Sweetgum 1 3 25 17 64 Sweetgum 1 2 19 17 65 Pine 1 4 34 17 66 Sweetgum 1 3 19 17 67 Pine 1 2 25 17 68 Oak 1 2 16 17 69 Pine 1 2 21 17 70 Pine 1 2 21 17 71 Sweetgum 1 4 32 17 72 Pine 1 6 43

University of Texas Isoprene Study ‐ January 2018

82

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

17 73 Sweetgum 1 3 19 17 74 Magnolia 1 2 21 17 75 Magnolia 1 1 16 17 76 Oak 1 3 21 17 77 Pine 1 6 46 17 78 Sweetgum 1 3 22 17 79 Pine 1 2 19 17 80 Pine 1 2 20 17 81 Sweetgum 1 3 26 17 82 Magnolia 1 2 21 17 83 Pine 1 2 18 17 84 Pine 1 4 35 17 85 Pine 1 4 35 17 86 Pine 1 3 27 17 87 Pine 1 4 41 17 88 Pine 1 4 36 17 89 Magnolia 1 3 23 17 90 Magnolia 1 4 26 17 91 Sweetgum 1 3 26 17 92 Sweetgum 1 3 26 17 93 Pine 1 2 23 17 94 Oak 1 3 24 17 95 Oak 1 1 19 17 96 Magnolia 1 5 36 18 1 Oak 1 2 15 18 2 Oak 1 14 16 18 3 Oak 1 3 23 18 4 Pine 1 1 26 18 5 Oak 1 8 42 18 6 Pine 1 2 23 18 7 Sweetgum 1 2 16 18 8 Oak 1 5 15 18 9 Pine 1 2 24 18 10 Oak 1 2 19 18 11 Oak 1 7 35 18 12 Oak 1 7 34 18 13 Magnolia 1 2 16 18 14 Oak 1 5 23 18 15 Oak 1 4 21 18 16 Oak 1 5 23 18 17 Redbay 1 2 18 18 18 Magnolia 1 3 26 18 19 Magnolia 1 4 31 18 20 Pine 1 4 25

University of Texas Isoprene Study ‐ January 2018

83

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

18 21 Magnolia 1 5 34 18 22 Pine 1 3 24 18 23 Oak 1 2 17 18 24 Oak 1 4 33 18 25 Pine 1 5 37 18 26 Pine 1 9 37 18 27 Pine 1 3 31 18 28 Pine 1 3 26 18 29 Pine 1 2 18 18 30 BlackGum 1 3 17 18 31 Holly 1 2 16 18 32 Pine 1 1 17 18 33 Sweetgum 1 6 29 18 34 Oak 1 5 25 18 35 Holly 1 2 19 18 36 Holly 1 3 18 18 37 Holly 1 4 26 18 38 Holly 1 7 34 18 39 Pine 1 2 17 18 40 Oak 1 7 36 18 41 Oak 1 4 31 18 42 Redbay 1 4 21 18 43 Pine 1 3 17 18 44 Pine 1 5 34 18 45 Oak 1 2 15 19 1 Magnolia 1 9 32 19 2 Magnolia 1 1 16 19 3 Magnolia 1 7 50 19 4 Magnolia 1 6 50 19 5 Magnolia 1 13 60 19 6 Magnolia 1 5 40 19 7 BlackGum 1 6 40 19 8 Oak 1 11 62 19 9 Sweetgum 1 4 25 19 10 Sweetgum 1 9 70 19 11 Sweetgum 1 11 75 19 12 BlackGum 1 2 15 19 13 BlackGum 1 4 34 19 14 BlackGum 1 3 30 19 15 BlackGum 1 4 36 19 16 Oak 1 15 70 20 1 Holly 1 7 41 20 2 Holly 1 1 15 20 3 Maple 1 8 46

University of Texas Isoprene Study ‐ January 2018

84

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

20 4 Oak 1 9 45 20 5 Oak 1 4 23 20 6 Oak 1 2 19 20 7 Oak 1 4 41 20 8 Oak 1 6 41 20 9 Oak 1 15 85 20 10 Holly 1 2 22 20 11 Holly 1 2 22 20 12 Holly 1 2 22 20 13 Holly 1 1 22 20 14 Holly 1 3 20 20 15 Holly 1 9 34 20 16 Oak 1 4 29 20 17 Pine 1 7 47 20 18 Dogwood 1 6 35 20 19 Dogwood 1 6 35 20 20 Holly 1 1 15 20 21 Holly 1 2 19 20 22 Sweetgum 1 10 61 20 23 Sweetgum 1 8 55 20 24 Oak 1 11 57 20 25 Oak 1 4 30 20 26 Oak 1 5 51 20 27 Oak 1 2 23 20 28 Magnolia 1 2 18 20 29 Oak 1 9 45 20 30 Oak 1 3 26 20 31 Oak 1 18 75 20 32 Magnolia 1 3 24 20 33 Oak 1 14 71 20 34 Holly 1 9 47 20 35 Holly 1 2 19 20 36 Holly 1 2 18 20 37 Holly 1 2 19 20 38 Pine 1 7 47 20 39 Oak 1 4 32 20 40 Oak 1 4 32 20 41 Holly 1 5 39 20 42 Holly 1 3 21 20 43 Holly 1 3 20 21 1 Pine 1 10 70 21 2 Pine 1 9 71 21 3 Sweetleaf 1 1 16 21 4 Sweetleaf 1 2 18

University of Texas Isoprene Study ‐ January 2018

85

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

21 5 BlackGum 1 7 42 21 6 Sweetleaf 1 2 21 21 7 Oak 1 4 34 21 8 Sweetleaf 1 2 19 21 9 Sweetleaf 1 3 31 21 10 Sweetleaf 1 3 34 21 11 Oak 1 14 47 21 12 Oak 1 14 53 21 13 Sweetleaf 1 3 32 21 14 Holly 1 5 43 21 15 Pine 1 2 28 21 16 Pine 1 2 27 21 17 Sweetleaf 1 2 16 21 18 BlackGum 1 9 53 21 19 Sweetleaf 1 4 28 21 20 Sweetgum 1 11 63 21 21 Sweetleaf 1 2 22 21 22 Pine 1 5 47 21 23 Redbay 1 1 19 21 24 Oak 1 9 49 21 25 Pine 1 13 63 21 26 Pine 1 13 65 21 27 Sweetleaf 1 2 30 21 28 Sweetleaf 1 4 32 21 29 Sweetleaf 1 4 33 21 30 Oak 1 8 56 21 31 Holly 1 6 43 21 32 Holly 1 2 23 21 33 Holly 1 2 19 21 34 Oak 1 6 40 21 35 Oak 1 17 50 21 36 Pine 1 9 42 21 37 Holly 1 4 31 21 38 Holly 1 6 37 21 39 Sweetleaf 1 2 19 21 40 Holly 1 8 48 22 1 Oak 1 4 23 22 2 Oak 1 4 26 22 3 Oak 1 2 22 22 4 Pine 1 11 47 22 5 Oak 1 4 27 22 6 Magnolia 1 1 17 22 7 Oak 1 1 17 22 8 Oak 1 3 31

University of Texas Isoprene Study ‐ January 2018

86

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

22 9 Oak 1 1 17 22 10 Oak 1 1 18 22 11 Oak 1 2 22 22 12 Oak 1 2 19 22 13 Pine 1 3 26 22 14 Magnolia 1 3 19 22 15 Magnolia 1 9 47 22 16 Sweetgum 1 12 52 22 17 Oak 1 4 22 22 18 Pine 1 7 37 22 19 Pine 1 2 23 22 20 Oak 1 2 22 22 21 Oak 1 1 17 22 22 Elm 1 2 18 22 23 Oak 1 3 21 22 24 Oak 1 2 20 22 25 Oak 1 1 18 22 26 Oak 1 2 17 22 27 Sweetleaf 1 1 17 22 28 Oak 1 2 22 22 29 Oak 1 3 32 22 30 Holly 1 3 23 22 31 Sparkleberry 1 1 17 22 32 Sparkleberry 1 2 17 22 33 Sparkleberry 1 2 17 22 34 Sparkleberry 1 1 17 22 35 Oak 1 24 53 22 36 Oak 1 7 36 22 37 Oak 1 2 19 22 38 Sweetgum 1 5 34 22 39 Oak 1 2 21 22 40 Oak 1 2 22 23 1 Oak 1 3 23 23 2 Oak 1 1 19 23 3 Oak 1 2 21 23 4 Oak 1 1 17 23 5 Oak 1 1 15 23 6 Oak 1 1 17 23 7 Oak 1 2 19 23 8 Pine 1 5 40 23 9 Sweetgum 1 8 50 23 10 Magnolia 1 4 41 23 11 Oak 1 6 36 23 12 Oak 1 4 33

University of Texas Isoprene Study ‐ January 2018

87

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

23 13 Magnolia 1 3 26 23 14 Pine 1 10 57 23 15 Oak 1 4 33 23 16 Sweetgum 1 3 25 23 17 Oak 1 4 37 23 18 Sweetgum 1 5 31 23 19 Sweetgum 1 5 29 23 20 Pine 1 2 21 23 21 Sweetgum 1 5 33 23 22 Sweetgum 1 2 17 23 23 Sweetgum 1 4 26 23 24 Pine 1 3 20 23 25 Pine 1 3 21 23 26 Sweetgum 1 5 27 23 27 Sweetgum 1 5 29 23 28 Maple 1 7 35 23 29 Maple 1 7 32 23 30 Pine 1 5 36 23 31 Pine 1 5 18 23 32 Maple 1 4 22 23 33 Pine 1 3 21 23 34 Sweetgum 1 4 24 23 35 Maple 1 3 27 23 36 Sweetgum 1 4 36 23 37 Pine 1 8 51 23 38 Sweetgum 1 3 26 23 39 Sweetgum 1 3 33 23 40 Sweetgum 1 5 37 23 41 Oak 1 15 65 23 42 Sweetgum 1 2 17 23 43 Sweetgum 1 4 30 23 44 Sweetgum 1 3 25 23 45 Sweetgum 1 3 30 23 46 Sweetgum 1 3 31 23 47 Oak 1 5 47 23 48 Sweetgum 1 4 34 23 49 Sweetgum 1 2 23 23 50 Sweetgum 1 6 37 23 51 Sweetgum 1 3 31 23 52 Sweetgum 1 2 25 23 53 Sweetgum 1 6 37 23 54 Sweetgum 1 4 19 23 55 BlackGum 1 4 19 23 56 BlackGum 1 3 19

University of Texas Isoprene Study ‐ January 2018

88

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

23 57 Pine 1 7 55 23 58 Pine 1 8 55 23 59 Pine 1 7 61 23 60 Oak 1 17 54 23 61 Oak 1 2 21 23 62 Pine 1 4 31 23 63 Sweetgum 1 8 35 23 64 Sweetgum 1 3 19 23 65 Oak 1 1 19 23 66 Oak 1 10 45 23 67 Sweetgum 1 5 34 24 1 Oak 1 4 24 24 2 BlackGum 1 4 21 24 3 Oak 1 1 19 24 4 Chinkapin 1 1 22 24 5 Oak 1 10 36 24 6 Pine 1 25 90 24 7 Oak 1 1 16 24 8 Oak 1 17 53 24 9 Oak 1 4 30 24 10 BlackGum 1 3 35 24 11 Oak 1 2 23 24 12 Oak 1 3 20 24 13 BlackGum 1 2 23 24 14 Oak 1 10 65 24 15 Oak 1 9 59 24 16 Sweetleaf 1 1 16 24 17 Sweetleaf 1 1 15 24 18 Oak 1 2 26 24 19 BlackGum 1 4 23 24 20 Oak 1 10 49 24 21 Oak 1 1 18 24 22 Oak 1 8 45 24 23 Magnolia 1 3 20 24 24 Oak 1 4 26 24 25 Oak 1 11 47 24 26 Oak 1 2 17 24 27 Oak 1 2 19 24 28 Sparkleberry 1 2 16 24 29 Sparkleberry 1 2 16 24 30 Oak 1 2 19 24 31 Oak 1 2 17 24 32 Redbay 1 1 16 24 33 Oak 1 1 16

University of Texas Isoprene Study ‐ January 2018

89

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

24 34 Oak 1 1 16 24 35 Oak 1 3 23 24 36 Oak 1 2 23 24 37 Oak 1 1 19 24 38 Oak 1 2 20 24 39 Oak 1 2 24 24 40 BlackGum 1 3 20 24 41 Oak 1 1 18 24 42 Oak 1 2 22 24 43 Oak 1 1 19 24 44 Oak 1 4 28 24 45 Oak 1 1 19 24 46 Oak 1 6 23 24 47 Holly 1 3 23 25 1 Holly 1 9 50 25 2 BlackGum 1 4 30 25 3 Pine 1 18 90 25 4 Pine 1 15 85 25 5 Oak 1 8 50 25 6 BlackGum 1 4 40 25 7 Sweetgum 1 3 17 25 8 Magnolia 1 2 21 25 9 BlackGum 1 1 15 25 10 Oak 1 11 52 25 11 Magnolia 1 3 34 25 12 BlackGum 1 3 23 25 13 Oak 1 10 45 25 14 Magnolia 1 2 19 25 15 Magnolia 1 1 16 25 16 Oak 1 3 25 25 17 Oak 1 4 17 25 18 Pine 1 26 98 25 19 Pine 1 13 87 25 20 Pine 1 13 84 25 21 Oak 1 7 37 25 22 Oak 1 6 37 25 23 Oak 1 8 59 25 24 Oak 1 6 37 25 25 Oak 1 6 33 25 26 BlackGum 1 5 35 25 27 Pine 1 11 60 25 28 Pine 1 10 55 25 29 Oak 1 7 37 25 30 Oak 1 4 17

University of Texas Isoprene Study ‐ January 2018

90

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

25 31 Pine 1 20 85 25 32 BlackGum 1 4 24 25 33 Oak 1 8 49 25 34 BlackGum 1 4 37 25 35 Oak 1 7 45 25 36 Oak 1 2 20 25 37 Magnolia 1 2 27 25 38 Oak 1 8 62 25 39 Oak 1 10 60 25 40 Pine 1 14 92 26 1 Oak 1 12 69 26 2 Hickory 1 12 77 26 3 Oak 1 5 39 26 4 Holly 1 3 27 26 5 Oak 1 16 65 26 6 Dogwood 1 3 24 26 7 Ironwood 1 4 23 26 8 Hickory 1 2 22 26 9 Holly 1 6 34 26 10 Ironwood 1 2 22 26 11 Oak 1 11 72 26 12 Maple 1 5 33 26 13 Oak 1 6 40 26 14 Holly 1 5 31 26 15 Hickory 1 12 55 26 16 Ash 1 14 58 26 17 Holly 1 9 45 26 18 Ironwood 1 5 40 26 19 Oak 1 2 24 26 20 Sweetgum 1 5 34 26 21 Oak 1 2 28 26 22 Beech 1 7 48 26 23 Ironwood 1 5 44 26 24 BlackGum 1 3 23 26 25 Oak 1 15 65 26 26 BlackGum 1 15 76 26 27 Ironwood 1 1 20 26 28 Holly 1 4 32 26 29 Holly 1 4 36 26 30 Oak 1 8 44 26 31 Oak 1 5 28 26 32 Oak 1 4 37 26 33 Oak 1 1 27 26 34 Ironwood 1 6 41

University of Texas Isoprene Study ‐ January 2018

91

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

26 35 Ironwood 1 5 37 27 1 Oak 1 9 42 27 2 Oak 1 9 45 27 3 Oak 1 7 31 27 4 Oak 1 2 25 27 5 Oak 1 1 17 27 6 Oak 1 2 18 27 7 Oak 1 2 19 27 8 Magnolia 1 1 23 27 9 Oak 1 1 17 27 10 Oak 1 3 26 27 11 Oak 1 1 21 27 12 Oak 1 1 20 27 13 Oak 1 1 18 27 14 Oak 1 1 18 27 15 Magnolia 1 1 18 27 16 Magnolia 1 2 18 27 17 Oak 1 3 19 27 18 Oak 1 5 28 27 19 Pine 1 12 62 27 20 Oak 1 11 56 27 21 Oak 1 3 20 27 22 Oak 1 2 21 27 23 Oak 1 2 15 27 24 Oak 1 6 29 27 25 Oak 1 5 27 27 26 Pine 1 3 23 27 27 Oak 1 2 17 27 28 Oak 1 2 18 27 29 Magnolia 1 2 17 27 30 Sweetgum 1 2 17 27 31 Pine 1 4 35 27 32 Pine 1 2 19 27 33 Oak 1 12 46 27 34 Oak 1 1 17 28 1 Oak 1 4 24 28 2 Oak 1 2 21 28 3 Pine 1 14 69 28 4 Magnolia 1 2 26 28 5 Pine 1 17 98 28 6 Oak 1 8 49 28 7 Sweetgum 1 7 48 28 8 Sweetgum 1 5 32 28 9 Sweetgum 1 9 56

92

Point Number

Number

Species

Tree Count

Diameter Breast Height

in inches

Total Height in Feet

28 10 Magnolia 1 6 53 28 11 Sweetgum 1 6 41 28 12 Sweetgum 1 3 15 28 13 Sweetgum 1 4 36 28 14 Holly 1 8 41 28 15 Magnolia 1 5 27 28 16 Sweetgum 1 9 61 28 17 Oak 1 3 26 28 18 Sweetgum 1 3 23 28 19 Oak 1 1 16 28 20 Sweetgum 1 6 45 28 21 Oak 1 10 37 28 22 Sweetgum 1 3 21 28 23 Sweetgum 1 4 21 28 24 Sweetgum 1 3 26 28 25 Sweetgum 1 12 55 28 26 Sweetgum 1 13 55 28 27 Magnolia 1 3 34 28 28 Oak 1 5 27 28 29 Oak 1 3 22 28 30 Oak 1 5 48 28 31 Oak 1 2 17 28 32 Sweetgum 1 6 48

TOTAL 1114

93

Total Tally by Tree Species Summary of 28 ‐ 1/10th Acre Plots Plot Radius - 37.2'

Species Tree Count Ash 1 Beech 1 BlackGum 83 Cherry Laurel 1 Chinkapin 1 Dogwood 8 Elm 1 Hickory 5 Holly 93 Ironwood 14 Magnolia 90 Maple 28 Oak 394 Pine 210 Redbay 11 Sparkleberry 8 Sumac 3 Sweetgum 139 Sweetleaf 23 Grand Total 1114

94

University of Texas Isoprene Study Big Thicket National Preserve Tyler County, Texas December 2017 through January 2018

All Stems Over 15 Feet High Measured and Tallied by DBH in Inches

and Total Height in Feet 28 - 1/10th Acre Circular Plots Randomly

Placed with .5 miles of a Monitoring Station

95

Table 7 Parker Forestry Consultants Understory Plot Summary

Plot #

Site Description

Understory Species

Underbrush Density

Underbrush Height

Estimate of Underbrush Stems

01 Black-Gum Titi Flat Titi & Privet Heavy Over head high 300+ 02 Hardwood-Pine-flat with evidence of flooding Titi Heavy Over head high 500+ 03 Hardwood Flat - no Pine Yaupon & Privet Medium Over head high 70 04 Hardwood-Pine - edge of swamp Titi & Privet Heavy Over head high 500+ 05 Hardwood-Pine - ridge on edge of flat Yaupon Medium Over head high 150 06 Hardwood-Pine - upland hardwood Yaupon Medium Over head high 120 07 Pine-Hardwood -scattered pine and tall Black Gum Titi & Privet Heavy Over head high 500+ 08 Pine-Hardwood - next to Titi-Gum flat Privet Heavy Knee High - burned 500+ 09 Hardwood-Pine - edge of flat, transition to ridge Yaupon Low Over head high 30 10 Hardwood-Pine - mature Titi Swamp Titi Low Well over head high 95 11 Hardwood-Pine - edge of flat with hdwd regeneration Titi Medium Over waist high 250

12 Hardwood-Pine Yaupon & Privet Medium Over waist high 130 13 Hardwood-Pine - transition out of Titi swamp into

Upland Yaupon, Privet & Titi Medium Ground to over head

high 180

14 Hardwood-Pine -oak with a few pines Yaupon & Privet Low Over head high 95 15 Hardwood-Pine - wet flat Titi Heavy Over head high 500+ 16 Pine-Hardwood - Longleaf Pine Ridge Yaupon & Titi Medium Over head high 150 17 Hardwood-Pine - ridge on edge of flat Yaupon Medium Over head high 150 18 Hardwood-Pine - hardwood-pine ridge Yaupon Medium Over head high 200 19 Hardwood Bottom - no pine Titi & Privet Heavy Over head high 300+ 20 Hardwood-Pine - upland hardwood Yaupon Medium Over head high 200 21 Hardwood-Pine - low,wet flat Titi & Privet Heavy Over head high 300+

22

Transition between flatland and pine

Yaupon, Oak Saplings, Privet & Sparkleberry

Heavy

Head High

300+

23 Hardwood-Pine - mixed hardwood-pine flat Titi Low Over head high 75 24 Hardwood - Edge of Titi Swamp Titi & Privet Medium Over head high 175 25 Pine-Hardwood - Next to Gum-Titi flat Yaupon & Privet Medium Over head high 90 26 Hardwood-Pine - many oaks and few pines Yaupon Medium Over head high 90

27

Hardwood-Pine - flatwoods with hardwood regeneration on a slight rise

Yaupon & Redbay

Heavy

Head high

350

28 Hardwood-Pine - intermittent drain Titi Medium Over head high 150

96

97

98