TECHNICAL MEMORANDUM
REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS
PREPARED FOR: Imperial Irrigation District
PREPARED BY: Formation Environmental
DATE: August 15, 2018
1 INTRODUCTION AND BACKGROUND Hourly meteorological and air quality data are collected by six permanent stations located around the Salton Sea at Bombay Beach, Naval Test Base, Salton City, Salton Sea Park, Sonny Bono, and Torrez Martinez. Maintenance and operation of these stations is currently conducted by Air Sciences Inc. (Air Sciences). After data collection, Air Sciences also completes a quality assurance and quality control (QAQC) review and conducts data validation. To assure high-quality work products, Formation Environmental (Formation) completes an additional review of the data prior to its release to the public. This memorandum describes: Formation’s supplementary data review process; comments and questions prepared by Formation on the 2016, 2017, 2018 first quarter, and 2018 second quarter datasets; and responses to comments on the aforementioned datsets provided by Air Sciences.
2 DATA REVIEW PROCESS The following section describes the primary and supplementary meteorological and air quality data review processes.
2.1 PRIMARY DATA REVIEW PROCESS Air Sciences’ data review and validation processes are described in the technical memorandum, Salton Sea – Ambient Air Quality Monitoring Network 2016-2017 Data Review, dated August 15, 2018. This document is provided as Attachment 1. Hereafter, it is referred to as the SSAQMN Tech Memo.
2.2 SUPPLEMENTARY DATA REVIEW PROCESS Prior to public release of the meteorological and air quality datasets, Formation conducts an internal supplementary review process, which consists of two phases:
• Screening of data relative to a series of criteria developed in consultation with the SSAQMN Tech Memo
1 REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS • Development and review of data through diagnostic plots
The data screening process and the development of the diagnostic plots were completed in R: A Language and Environment for Statistical Computing (hereafter referred to as R). The two phases of the supplementary review process are discussed in the following subsections.
2.2.1 DATA SCREENING PROCESS Each of the meteorological and air quality datasets are independently screened after the initial development and data QAQC process. The following table summarizes the supplementary data screens on a parameter basis. Values meeting the criteria specified below are flagged for further review.
TABLE 1. SSAQMN SUPPLEMENTARY DATA REVIEW Parameter Independent Data Screens Barometric Pressure - Values less than 0.9 atm - Values greater than 1.1 atm PM Concentrations - Values (PMC, PM2.5, PM10 local, and PM10 STP) less than -15 ug/cubic meter - Values for which the sum of PM2.5 and PM10 is not equal to PMC (within + 0.1 ug/cubic meter) Precipitation - Values less than 0 mm - Values greater than 10 mm Relative Humidity - Values less than 0% - Values greater than 100% Temperature - Values screened against monthly minimum temperatures - Values screened against monthly maximum temperatures Wind Direction - Values less than 0 degrees - Values greater than 360 degrees - Values that are constant for 4 or more hours Wind Direction Standard Deviation - Values less than 0 degrees - Values greater than 100 degrees Wind Gust (maximum measured at - Values less than 0 m/s 10 m) - Values greater than 20 m/s - Values that are constant for 4 or more consecutive hours Wind Speed (horizontal) - Values less than 0 m/s - Values greater than 20 m/s - Values that are constant for 4 or more consecutive hours
Secondarily, the following situations are identified:
• Days for which the maximum temperature occurs prior to 9 a.m. • Times for which the temperature measured at 10 m minus the temperature measured at 2 m are positive during the day (indicating that the 10 m temperature exceeds the surface [2 m] temperature) or negative at night (indicating that the surface [2 m] temperature exceeds the 10 m temperature)
2 REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS • Days for which the maximum solar radiation does not occur mid-day (from 11 a.m. to 2 p.m.) • Times when the PM10 concentrations are less than the PM2.5 concentrations
These situations may occur due to meteorological conditions and are of lesser importance than the screening criteria outlined in Table 1. Unless other considerations indicate the need for further review, these situations are documented but not flagged for further review.
An additional data screen determines whether the Row Has Invalid field is properly completed. The Invalid Reasons field is also reviewed to confirm that the comments apply only to fields provided in the datasets.
2.2.2 REVIEW OF DIAGNOSTIC PLOTS For the second phase of the supplementary review process, the following diagnostic plots are developed and reviewed:
• Time series plots by parameter for all stations • Time series plots by parameter for each individual station
Additionally, wind roses are created from the 10 m wind speed data.
Concurrent review of the by-parameter time series plots for all stations and by individual station allows for inconsistencies between stations and potential issues at individual stations to be readily identified. Wind roses are reviewed for consistency with known prevailing wind speed and directionality patterns at each station.
3 SUMMARY OF COMMENTS AND RESPONSES
The following subsections provide a listing of the raw comments and questions provided to Air Sciences by Formation on each of the three datasets: 2016 and 2017 all quarters, 2018 first quarter, and 2018 second quarter.
3.1 2016 AND 2017 DATASETS The following comments and questions were prepared by Formation on the 2016 and 2017 datasets. Responses provided by Air Sciences follow each comment.
1. There are relative humidity measurements at 2m during 2016 Q1 and Q3 at Bombay Beach, 2016 Q1 and Q2 at Salton Sea Park, and 2016 Q1, Q2, and Q3 at Torres Martinez which are greater than 100%. According to the QAPP, "the humidity range of the humidity sensor is from 0-100 percent with an accuracy of ±0.8 percent when the temperature is at 23°C." Please clarify why these measurements are included in the final dataset.
3 REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS Response: “The stated accuracy for the RH probe is correct. However, the instruments do drift as they age (up to 1% per year). In addition, the EPA audit tolerance is 8%, so a drift below that will not be flagged for recalibration. The current logger program forces the maximum to 100% so this won't come up again, but the decision was made to leave values >100% for older data based on the above.”
2. There are ambient relative humidity measurements during 2016 Q3 at Naval Test Base which are below 0%. Please clarify why these measurements are included in the final dataset.
Response: “From our review it appears this only affects the RH sensor on the TEOM, not the 2m Temp/RH probe. Since the TEOM RH is not reported with the ‘public’ TEOM dataset, this was not prioritized for review/flagging of older data.”
3. The sonic 2D vector average wind direction measurements during 2016 Q1 at Sonny Bono do not span full range of 0-360 degrees. Wind direction measurements from other time frames at this station show wind directions from the full range. Please review and clarify.
Response: “You are correct and this should have been invalidated. We had a miscommunication with invalidating sonic data at SB. This has been corrected.”
4. Similarly, the wind direction measurements at 10m during 2016 Q1 at Sonny Bono do not span full range of 0-360 degrees. Please review and clarify. Additionally, please comment on the gap in collection (i.e. 2016 Q2 and 2016 Q3 data is missing).
Response: “From Tim: Given the full sonic invalidation at SB for this time period, and given no other information available to us to invalidate the WD_10m (vane) data – we opted to keep the SB data for the first part of 2016. There isn’t a full compass span for WD_10m for those limited ~45 days, but we have nothing as a basis to flunk the data…the CARB audit on 2/24/2016 included WD and indicated passing for all WD checks so we assumed that the 1/1 to 2/24 was good.
Wind direction failed the close-out audit at SB on 10/3/2016 and so we had to invalidate back to the last ‘known good’ point, which was the CARB audit on 2/24.”
5. When comparing the wind speed measurements at 1m and 2m, it seems that some instruments have a different lower cutoff value to reflect calm or still conditions. Is this correct?
Response: “Yes, the minimum threshold was set inconsistently in the Ecotech system. We did not attempt to standardize the minimums for older data, but the current logger program is consistent for 10m WS at all sites.”
6. The PM2.5 filter load during 2017 Q3 at Torres Martinez appears anomalous (i.e. inconsistent with other data). Please review and clarify.
Response: “Yes, this should have been flagged and is now fixed.”
4 REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS
7. The PM2.5 mass accumulations are negative during 2017 Q3 at Salton City and 2017 Q1, Q2 and Q3 at Salton Sea Park. Please review and clarify.
Response: “This correlates with brief (1-hour) periods where the TEOM calculates a negative concentration due to an instability in the instrument. We don't think it is an issue in regards to the concentration data being invalid (other than the brief period of extreme negative conc.), but we are getting clarification from Thermo on why this happens and if it's a problem. High relative humidity can contribute, and the example of Salton City in 2017 corresponds to RH peaks: basically if the incoming PM is hydrated, sticks to the filter, and then dries after a brief period it will destabilize the calcs for that hour, and since the calculation is simply a subtraction from the previous hour, the instrument can end up with negative mass (this relative calculation is why the error doesn't persist in subsequent hours for concentration, but it ends up affecting mass accumulation for all subsequent hours because the values are additive). We will follow up once we have a definitive answer.”
8. The PM2.5 mass transducer oscillator noise is essentially zero for most of 2016 Q1 and Q2 at Bombay Beach, Sonny Bono, and Torres Martinez. The data appears distinctly different than other data. Please review and clarify.
Response: “This is most likely due to incorrect reporting by the Ecotech system and we didn’t prioritize it for individual flagging during this time period in our review, but it was used as a diagnostic for validating TEOM data if there was reason to believe that the concentration data were suspect.”
9. The vacuum bypass flows during 2017 Q2 at Naval Test Base, 2016 Q1 at Sonny Bono, and 2017 Q3 at Torres Martinez appear to be anomalous (i.e. inconsistent with other data). Please review and clarify.
Response: “The Q1 2016 values at Sonny Bono are from a data mapping problem from the Ecotech files: 144 is the status code and not the actual data. Similar to the oscillator noise, this parameter was not scrutinized individually as part of validating the TEOM concentrations for the older data and was left ‘as- is.’
The period of ‘jitter’ in 2017 at NTB corresponds to a period with a number of errors and site visits, including an audit where the flows were adjusted, but is within tolerance (12 l/min +/- 5%).
The anomalous period at TM corresponds with the PM2.5 filter load issue above; all of this data is now invalid.”
10. The vacuum pressure during 2016 Q1 and Q2 at Bombay Beach, Sonny Bono, Salton City, and Torres Martinez appear to be anomalous (i.e. inconsistent with other data). Please review and clarify.
Response: “We believe this is due to a rounding issue with data coming from the Ecotech system.”
11. There are multiple stations and dates for which the QAQC Level is set to “Not QC’ed”, “Level 1” or blank. Data should not be set to “Not QC’ed” or “Level 1”. Additionally, for stations and dates in
5 REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS which all data is invalid, sometimes the QAQC Level is “Level 2” and sometimes it is blank. Please choose a consistent approach.
Response: “Thank you for pointing this out, it revealed a subtle bug with our QA upgrade process that was dropping hours at the beginning or end of a QA'd period in rare circumstances. In addition, we didn't have a set way to identify the QA status for missing data (the blank rows) during the export process. This has all been fixed.”
12. Additionally, we need to address the issues with PM10-PM2.5-PMCoarse that Earl identified.
Response: “We are working on this now.”
13. We also had questions about the shelter temperatures at miscellaneous stations and dates, but your prior response to a question on the 2018 Q1 datasets was helpful in understanding those data.
Response: None.
The following comments and questions were prepared by Formation in August 2018 on revisions to the final 2016 and 2017 datasets. Responses provided by Air Sciences follow each comment. Communications were between Bronwyn Paxton (Formation) and Matt Mavko (Air Sciences).
14. Precipitation and barometric pressure are missing from the met and combined datasets.
Response: “As discussed on the phone, this was inadvertently left out.”
15. There are many cases in the TEOM private and public datasets where the Row Has Invalid Flag is not consistent with the dataset (i.e. NA’s exist in the dataset and the Row Has Invalid is equal to false).
Response: “These were isolated to a period Oct‐Dec 2016 that I forgot to refresh. This is part of the same phenomenon that I fixed for 2018 TM data previously where STP values flipped from NA to valid. I did not refresh this period, forgetting we switched from the old system in late 2016. I refreshed the STP calcs for this period and the NA values without flags are now gone. I checked 2017 and 2018 and these periods are both clean.”
16. The Sonny Bono data (Ambient Temperature at 2m, Relative Humidity at 2m, wind direction at 10m, wind speed at 10m, wind speed at 2m, and wind speed at 1m) from 2016‐01‐01 01:00:00 were numeric values in the previous version of the final dataset and then were converted to NA in this version. Please help us understand why these values have changed from the previous final dataset to the current dataset.
Response: “This inadvertently happened when separating a flag for reported vs unreported data, and the ending timestamp ended up off by one hour. We confirmed that there is valid data for this hour and adjusted the relevant flag back by an hour, so this will show valid data again.”
6 REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS 17. There are numerous data in the TEOM private, TEOM public, and combined datasets that differ from the previous final dataset (either from value to NA or NA to a value). Please see the attached spreadsheet for a summary of these discrepancies. Please help us understand why these values have changed from the previous final dataset to the current dataset.
Response: “For the values that went from a value to NA o Sonny Bono 1/1/16 met ‐ See above o Salton City 3/10/17 PM coarse ‐ cleaned up flag to remove unreported parameters; in the process, noticed that coarse PM was not flagged originally due to analyst error. Invalidating the entire table instead of selecting individual fields corrected this o TM 8/4/17 13:00 ‐ similar to SC above, cleaning up flag parameters corrected a PM table flag o SC 10/4/16 21:00 ‐ this ended up invalidated as part of identifying periods where PM10 stp was simply missing in the old system...this was during the TEOM transition period at the station and a flag was corrected, so NA is correct.
Values that changed from NA to a value, only PM 10 STP was affected and was due to the "refresh" process mentioned above and in previous email regarding TM.”
18. Please adjust the following Invalid Reason: • \"Invalidated Data ‐ (Coarse PM, PM10 QA Check): Data invalidated manually by an analyst\" to remove mention of the PM10 QA Check field
Response: “Corrected invalidation description, it slipped through during clean up of 2016 reported vs non‐reported parameters”
19. There are many Invalid Reasons in the met and combined datasets which contain mentions of parameters not included in dataset including: Sonic wind component (vertical), Sonic wind component (N‐S), Sonic wind component (E‐W), Sonic speed of sound, Sonic temperature, Sonic wind elevation, Sonic error code, Sonic Azimuth Wind Direction, Sonic 3D Sigma Theta, Sonic 3D vector‐avg. wind‐speed, Sonic 2D Sigma Theta, Sonic 3D vector‐avg. wind direction, Sonic 2D wind speed, Sonic 3D wind speed, U wind vector, V wind vector etc.
Response: “The sonic flags ‐ this is similar to wanting to flag the TEOM diagnostic fields even though not reported. The sonic has many parameters and by flagging all of them we are showing that all sonic data was bad. It is not a subjective thing like the TEOM. But, as we agreed on the phone, I will pull them to avoid confusion.”
20. There are a number of cases in the TEOM private datasets in which the Row Has Invalid flag is set to false yet there are numerous fields that are NA.
Response: Per a phone discussion on August 10, 2018, Matt indicated that there are a number of cases in which TEOM data had to be manually downloaded from the TEOM. The parameters stored internally in the TEOM are not identical to the parameters transmitted across the network. Therefore, the values were not invalid but instead considered missing. Matt will develop a README File.
7 REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS 21. In the met datasets, Salton City (2016‐10‐04 21:00) the ambient temperature at 2m, net calculated radiation, relative humidity at 2m, wind speed at 10m, wind speed at 2m, and wind speed at 1m were values and now are NA.
Response: Per a phone discussion with Matt Mavko on August 10, 2018, Salton City 2016‐10‐04 21:00 was right in the middle of the station transition. Matt realized that data before and after was invalidated and determined that the entire period should have been invalidated.
22. At Sonny Bono on 2016‐01‐01 01:00, the ambient relative humidity, ambient temperature, coarse PM, and TEOM operation mode error code were values and are now NA (with the exception of coarse PM which was NA and then became a value).
Response: Per a phone discussion with Matt Mavko on August 10, 2018, the conversion of the ambient relative humidity, ambient temperature, and TEOM operation to NAs was a mistake. Matt will rerun the 2016 Q1 Sonny Bono TEOM private dataset.
23. At Naval Test Base on 2016‐03‐29 01:00, the PM10 QA check went from NA to a value.
Response: Per a phone discussion with Matt Mavko on August 10, 2018, this was a correction associated with the QA Check comment. A QA Check value should not be invalidated (as it is a calculation) unless the raw values were invalidated.
3.2 2018 FIRST QUARTER DATASETS The following comments and questions were prepared by Formation on the 2018 first quarter datasets. Responses provided by Air Sciences follow each comment.
1. Why is the 10 m temperature data not included?
Response: “Temperature at 10m is part of a pair of sensors for performing the Delta Temperature (DT) calculation. The individual temperatures for those probes are not reported, only the calculated DT is reported. We do it this way because reporting one-hour temperature values for these probes can lead to confusion as it may seem like 10 minus 2 should equal DT, but since DT is calculated at the logger scan rate, it won’t match. Calculated DT from the one-hour temperature values is not a valid method. There is an independent temp/RH probe at 2m; these are reported in the met dataset.”
2. The shelter temperatures at Bombay Beach and Sonny Bono are substantially different from the other stations during January (as is the comparison with ambient temperature; see plots below). Can you explain this?
8 REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS
Response: “The shelter temperature readings at the AQ stations, in general, are terrible. We inherited these little thermisters from the previous operators that attach to the backs of the TEOMs, so the values are affected by heat from the TEOMs. We use them only as a relative diagnostic check to indicate if the AC is failing, which has worked well enough so far (it’s pretty obvious when the values start tracking ambient). In January, Ramon started making adjustments to the linear conversion equation to see if it
9 REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS helped the temps match better the wall thermometer for the AC unit. However, this activity wasn’t mentioned in the site visit forms so we're not sure exactly when the adjustments were made/undone. Installing better thermometers would be time-consuming and expensive so we haven't made it a priority.”
3. Why is the PM2.5 filter load at Torres Martinez substantially different than the others from mid- January to early February (see plots below)? It seems like the PMC filter load was invalidated during the same time frame.
10 REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS
Response: “The pm2.5 filter loading calculation by the TEOM was affected by a site visit where Ramon replaced only the PMC filter and did not touch the pm2.5 filter. The internal calcs of the TEOM subsequently got scrambled, and a subsequent site visit corrected the issue. The mass accumulation measurement and oscillation frequency of the mass transducer were unaffected so we have no reason to believe that the PM2.5 conc values were invalid during this period. However, the filter loading values during this period are technically incorrect and could be invalidated. However, since loading is calculated indirectly by the TEOM by making certain assumptions about its operational state, it isn’t a priority to scrutinize heavily, we just use it as one diagnostic indicator that the filters need to be changed.”
4. Please clarify why the PM coarse vacuum flow at Bombay Beach, Sonny Bono, and Torres Martinez is scattered during isolated time frames and relatively consistent during others.
Response: “There are two related flow parameters that can be exported from the TEOMs: mass flow and volumetric flow (volumetric is corrected and is therefore more stable). For periods when we had to backfill data directly from the TEOM, PMC mass flow was substituted for PMC volumetric flow at these three sites and caused the data to be noisier. We are going to make the data exports from both methods consistent so that this confusion does not arise again.”
5. Please clarify why the vacuum bypass flow at Sonny Bono and Torres Martinez is scattered during isolated time frames and relatively consistent during others.
Response: “Same answer and corrective action as above: mass vs. volumetric flow for backfilled periods.”
11 REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS 6. For stations and dates in which all data is invalid, sometimes the QAQC Level is “Level 2” and sometimes it is blank. Please choose a consistent approach.
Response: “Yes, see answer for related comment above.”
The following comments and questions were prepared by Formation in August 2018 on revisions to the final 2018 Q1 datasets. Responses provided by Air Sciences follow each comment. Communications were between Bronwyn Paxton (Formation) and Matt Mavko (Air Sciences).
7. In the TEOM public, TEOM private, and combined datasets, there are 57 records in 2018 Q1 and Q2 (see attached list of deployments and dates) for which PM10 at STP differs between previous version and updated version (it is NA in the previous dataset; there is a value in the updated dataset). Would you take a look and let us know why these discrepancies are occurring?
Response: “I noticed in reviewing the TM data today that there were some stray STP PM10 values that had not calculated correctly the first time so I refreshed those. They were not flagged as invalid, either, (i.e. NA with no invalid reason) so they needed to be corrected. Sorry for not mentioning.”
8. In the Invalid Reason field of the combined datasets, there are mentions of the U and V wind vectors. Please remove these mentions.
Response: “The U and V vector mentions were removed (only affected NTB). Our analyst included them so that they were correct for modeling export, but I agree it’s confusing to include when we aren’t reporting them. All the parameters were combined into one flag and I separated them out so they won’t show up.”
9. In the meteorological dataset, at the Naval Test Base on 2018‐01‐04 17:00, the record is flagged as Row Has Invalid = FALSE despite the Wind Speed Max Gust at 10 m/s being set to NA. Would you take a look and let us know why these discrepancies are occurring?
Response: “NTB wind_gust value on 1/4 at 17:00 is missing, and no flag was created. I added a flag for now to make it clear but in the future I will add a check for missing parameters to add a comment even if there is no flag.”
10. In the 2018 Q1 Torres Martinez teom public and combined Q1‐Q2 datasets, there are two flags which need to be adjusted to exclude mentions of data fields which are not available: • "[\"Invalidated Data ‐ (PM10 STP, PM2.5 Vacuum Flow, Vacuum Bypass Flow, PM2.5 Mass Accumulation, PMC Mass Accumulation, TEOM Dew Point): Data invalidated manually by an analyst\",\"Invalidated Data ‐ (All Fields): Data invalidated manually by an analyst\"]" • "[\"Invalidated Data ‐ (PM10 STP, PM2.5 Vacuum Flow, Vacuum Bypass Flow, PM2.5 Mass Accumulation, PMC Mass Accumulation, TEOM Dew Point): Data invalidated manually by an analyst\"]"
12 REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS Response: These flags were initially left alone because they were helpful in understanding the dataset invalidations. However, after further discussion, these flags were removed from the public datasets so as not to create additional confusion by public users.
11. There are 65 changes in TEOM private and public datasets (PM10 STP) compared to the previous version. • 51 changes were noted in previous comments and discussed • It appears that there are 14 additional records from Torres Martinez that were changed from NA to a value (1/23/2018 8:00:00 PM to 1/24/2018 9:00:00 AM) from the 3 previous final dataset to this version. Please help us understand why these values have changed from the previous final dataset to the current dataset.
Response: “I had removed an erroneous flag that was invalidating just PM10 STP from 1/23 ‐ 1/24, causing the change.”
3.3 2018 SECOND QUARTER DATASETS The following comments and questions were prepared by Formation on the 2018 second quarter datasets. Responses provided by Air Sciences follow each comment.
1. At Sonny Bono, from mid-April to mid-May, the calculated temperature difference between 10m and 2m as well as the maximum gust wind speed were identified as "invalidated manually by an analyst." Please describe the underlying issues.
Response: “Back in early 2017, we removed the CR1000 at SB to send it to Campbell for calibration and installed a CR3000 from the Air Sciences audit kit. Katie finally reinstalled the CR1000 in April 2018, and wired the delta temperature probes backward (i.e., swapped probe heights; she checked them real-time during the swap and they seemed reasonable, but because it was mid-day there wasn’t much of a difference between the two and the problem went undetected). Upon reinstalling the CR1000, we lost the real-time data feed to the logger and could not get it fixed until Kyle returned from leave, which was in mid-May. Thus, there was a delay in catching the wiring error at the station. We corrected the issue in mid-May. We will add more detail to the invalidation comment.”
2. At Sonny Bono, in mid-June, the variability and trends in vacuum pressure differ from the rest of the time period (shown below). Please clarify.
13 REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS
Response: “Unfortunately, we're not really sure why this happened. However, the flow rates during this period were unaffected, which is what directly affects the concentration validity (the TEOM can compensate for variations in pressure). In addition, despite being more variable, it is still within the valid range.”
3. At Salton City, on 05/01/2018 (beginning at 14:00) there is a six hour period where PM10_QA_Check_ugpcm is elevated. Please review and clarify.
Response: “The difference is within the expected range during periods of increasing filter loading/elevated concentrations (Coarse PM peaked at 1,600 µg/m^3 during this period, and was > 600 for seven hours straight).”
The following comments and questions were prepared by Formation in August 2018 on revisions to the final 2018 Q2 datasets. Responses provided by Air Sciences follow each comment. Communications were between Bronwyn Paxton (Formation) and Matt Mavko (Air Sciences).
4. In the TEOM public, TEOM private, and combined datasets, there are 57 records in 2018 Q1 and Q2 (see attached list of deployments and dates) for which PM10 at STP differs between previous version and updated version (it is NA in the previous dataset; there is a value in the updated dataset). Would you take a look and let us know why these discrepancies are occurring?
14 REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS Response: “I noticed in reviewing the TM data today that there were some stray STP PM10 values that had not calculated correctly the first time so I refreshed those. They were not flagged as invalid, either, (i.e. NA with no invalid reason) so they needed to be corrected. Sorry for not mentioning.”
5. In the TEOM public and TEOM private datasets, for the Bombay Beach 2018 Q2 datasets, both previous and updated versions of the datasets are being flagged as Row Has Invalid = FALSE and have no Invalid Reasons despite NA’s being present in the previous version of the dataset.
Response: “I don’t see what you are seeing for Bombay Beach. My files are clean. I am worried that Google Drive didn’t refresh in time and you have an older version? I just pulled the BB files to my home computer and they do not have any NA’s without “Has Invalid”=TRUE. I attached them here so you can check.
There were two records in Salton City where the flag reason was “null” because the flag didn’t have a description. Corrected.”
6. In the Invalid Reason field of the combined datasets, there are mentions of the U and V wind vectors. Please remove these mentions.
Response: “The U and V vector mentions were removed (only affected NTB). Our analyst included them so that they were correct for modeling export, but I agree it’s confusing to include when we aren’t reporting them. All the parameters were combined into one flag and I separated them out so they won’t show up.”
7. In the 2018 Q1 Torres Martinez teom_public and combined Q1‐Q2 datasets, there are two flags which need to be adjusted to exclude mentions of data fields which are not available: • "[\"Invalidated Data ‐ (PM10 STP, PM2.5 Vacuum Flow, Vacuum Bypass Flow, PM2.5 Mass Accumulation, PMC Mass Accumulation, TEOM Dew Point): Data invalidated manually by an analyst\",\"Invalidated Data ‐ (All Fields): Data invalidated manually by an analyst\"]" • "[\"Invalidated Data ‐ (PM10 STP, PM2.5 Vacuum Flow, Vacuum Bypass Flow, PM2.5 Mass Accumulation, PMC Mass Accumulation, TEOM Dew Point): Data invalidated manually by an analyst\"]"
Response: These flags were initially left alone because they were helpful in understanding the dataset invalidations. However, after further discussion, these flags were removed from the public datasets so as not to create additional confusion by public users.
8. There are 83 changes in combined dataset (PM10 STP) • 6 occurred during 2018 Q2 (consistent with previous comments and discussion) • In the Torres Martinez combined dataset, the first quarter data is duplicated (i.e. two identical rows for each date and time)
Response: “The duplicate rows and persistence of the old flag descriptions in the combined file were due to having an old test dataset from the previous round of TM fixes. I removed it from the quarterly folder and re‐ran the combiner script to fix.”
15 REVIEW PROCESS FOR 2016, 2017, 2018 FIRST QUARTER, AND 2018 SECOND QUARTER METEOROLOGICAL AND AIR QUALITY DATASETS 9. Per our phone discussion yesterday afternoon, the following Invalid Reasons were removed from the TEOM private dataset. However, they are still in the combined dataset: • "[\"Invalidated Data ‐ (PM10 STP, PM2.5 Vacuum Flow, Vacuum Bypass Flow, PM2.5 Mass Accumulation, PMC Mass Accumulation, TEOM Dew Point): Data invalidated manually by an analyst\",\"Invalidated Data ‐ (All Fields): Data invalidated manually by an analyst\"]" • "[\"Invalidated Data ‐ (PM10 STP, PM2.5 Vacuum Flow, Vacuum Bypass Flow, PM2.5 Mass Accumulation, PMC Mass Accumulation, TEOM Dew Point): Data invalidated manually by an analyst\"]"
Response: The PM comments were cleaned up.
4 ATTACHMENT The following item is attached to this memorandum:
• Attachment 1: Salton Sea – Ambient Air Quality Monitoring Network 2016-2017 Data Review prepared by Air Sciences and dated August 15, 2018
16 ATTACHMENT 1 ATTACHMENT 1
17
TECHNICAL MEMORANDUM
SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW
PREPARED FOR: Imperial Irrigation District
PREPARED BY: Formation Environmental, Air Sciences Inc., PlanTierra LLC
DATE: August 15, 2018
This technical memorandum provides an overview of data collected by the Salton Sea Air Quality Monitoring Network (SSAQMN) operated by the Salton Sea Air Quality Team (Team) for the period 2016-2017 and includes an overview of the data review QA/QC procedures used to review and validate the data gathered from the monitoring network.
This memo accompanies a data package that includes meteorological and particulate data for each station over the two-year period. Within the datasets, invalidated parameters are marked “NA” in place of a data value, and the documented reason(s) for the invalidation are concatenated into a single column in each file.
1 2016 DATA SUMMARY Major notable events in 2016 included the takeover of maintenance of the network by the Salton Sea Air Quality Team on August 1, 2016 and a reconfiguration of the stations starting in October 2016 and ending January 1, 2017. The station reconfiguration consisted of changing data-logging methods, removing the sonic anemometers, removing the 1m and 2m anemometers, and replacing old/outdated sensors. The sonic anemometers were plagued by calibration and maintenance problems and were deemed impractical to continue using. The 1m and 2m anemometers at each station were obstructed by fencing, and therefore were considered inadequately sited with no possible alternative for improvement.
1.1 2016 METEOROLOGICAL DATA HIGHLIGHTS Basic summary statistics are presented below for each of the six meteorological stations that are part of the SSAQMN. The close-out audits during the station reconfigurations resulted in several failed meteorological sensors that were consequently invalidated back to the previous audit (the CARB audit in 2016 Q1), resulting in low data recovery for some parameters. The net radiation sensors at several sites were found to be out of calibration during the 2016 Q1 CARB audit, and it was determined infeasible to bring them back into calibration after multiple attempts. These sensors were not replaced until the end of 2017.
1 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW
TABLE 1. BOMBAY BEACH METEOROLOGICAL DATA STATISTICS FOR 2016
Parameter Mean Max Min % Recovery Notes
Ambient Temperature 17.36 34.45 3.4 36.8% Failed close-out audit, invalidated (2m, °C) back to previous audit; sensor replaced 10/4/16 Barometric Pressure NA NA NA NA Not measured at this station (2m, atm) Delta Temperature 0.99 6.12 -0.96 22.0% Failed close-out audit, invalidated (°C) back to previous audit; sensors replaced 10/4/16 Net Radiation 16.11 335.3 -92.7 11.0% Failed CARB audit and was not able (Wm-2) to be recalibrated Precipitation NA NA NA NA Not measured at this station (mm) Relative Humidity 42.74 102.6 5.32 92.1% Sensor replaced 10/4/16 (2m, %) Solar Radiation NA NA NA NA Installed Dec 2017 (Wm-2) Sonic 2D WD 360 0 69.7% Frequent errors reported by (10m, deg) instrument and multiple missing data periods; Removed 10/4/16 Sonic 2D WS 3.29 18.5 0.06 69.7% Frequent errors reported by (10m, ms-1) instrument and multiple missing data periods; Removed 10/4/16 Wind Direction 360 0.03 22.0% Installed 10/4/16 (10m, deg) Wind Direction SD 23.01 100 2.85 22.0% Installed 10/4/16 (10m, deg) Wind speed 3.29 18.4 0.11 92.2% Sensor replaced 10/4/16 (10m, ms-1) Wind speed 2.26 13.2 0.2 70.2% Removed 10/4/16 (1m, ms-1) Wind speed 2.67 14.8 0.27 70.2% Removed 10/4/16 (2m, ms-1)
2 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW
TABLE 2. NAVAL TEST BASE METEOROLOGICAL DATA STATISTICS FOR 2016
Parameter Mean Max Min % Recovery Notes
Ambient Temperature 17.01 34.18 2.57 31.3% Failed close-out audit, invalidated (2m, °C) back to previous audit; sensor replaced 10/14/16 Barometric Pressure NA NA NA NA Not measured at this station (2m, atm) Delta Temperature 1.53 8.55 -0.92 17.6% Failed close-out audit, invalidated (°C) back to previous audit; sensors replaced 10/14/16 Net Radiation 0.0% Failed CARB audit and was not able (Wm-2) to be recalibrated Precipitation NA NA NA NA Not measured at this station (mm) Relative Humidity 32.75 99.4 0.3 88.9% Sensor replaced 10/14/16 (2m, %) Solar Radiation NA NA NA NA Installed Dec 2017 (Wm-2) Sonic 2D WD 359.91 0.12 49.5% Frequent errors reported by (10m, deg) instrument and multiple missing data periods; Removed 10/14/16 Sonic 2D WS 4.16 18.2 0.01 49.5% Frequent errors reported by (10m, ms-1) instrument and multiple missing data periods; Removed 10/14/16 Wind Direction 359.7 0.04 17.6% Installed 10/14/16 (10m, deg) Wind Direction SD 23.65 98.1 1.46 17.6% Installed 10/14/16 (10m, deg) Wind speed 4.29 18.44 0.07 89.0% Sensor replaced 10/14/16 (10m, ms-1) Wind speed 2.32 11.49 0.01 60.6% Removed 10/14/16 (1m, ms-1) Wind speed 2.96 12.96 0.01 68.6% Removed 10/14/16 (2m, ms-1)
3 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW
TABLE 3. SALTON CITY METEOROLOGICAL DATA STATISTICS FOR 2016
Parameter Mean Max Min % Recovery Notes
Ambient Temperature 24.62 47.95 0.42 82.3% Sensor replaced 10/14/16 (2m, °C) Barometric Pressure NA NA NA NA Not measured at this station (2m, atm) Delta Temperature 1.04 6.73 -1.25 21.7% Failed close-out audit, invalidated (°C) back to previous audit; sensors replaced 10/14/16 Net Radiation 52.42 487.34 -176.45 82.3% (Wm-2) Precipitation NA NA NA NA Not measured at this station (mm) Relative Humidity 36.97 100.1 4.3 82.3% Sensor replaced 10/14/16 (2m, %) Solar Radiation NA NA NA NA Installed Dec 2017 (Wm-2) Sonic 2D WD 359.7 0 39.6% Frequent errors reported by (10m, deg) instrument and multiple missing data periods; Removed 10/14/16 Sonic 2D WS 3.41 16.3 0 39.6% Frequent errors reported by (10m, ms-1) instrument and multiple missing data periods; Removed 10/14/16 Wind Direction 359.8 0.04 21.7% Installed 10/14/16 (10m, deg) Wind Direction SD 21.19 98.4 3.12 21.7% Installed 10/14/16 (10m, deg) Wind speed 3.54 16.9 0.19 82.3% Sensor replaced 10/14/16 (10m, ms-1) Wind speed 1.59 7.9 0.08 25.8% Removed 10/14/16 (1m, ms-1) Wind speed 2.13 9.79 0.11 25.8% Removed 10/14/16 (2m, ms-1)
4 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW
TABLE 4. SALTON SEA PARK METEOROLOGICAL DATA STATISTICS FOR 2016
Parameter Mean Max Min % Recovery Notes
Ambient Temperature 16.04 33.7 3.5 14.9% Failed close-out audit, invalidated (2m, °C) back to previous audit; sensor replaced 12/21/16 Barometric Pressure NA NA NA NA Not measured at this station (2m, atm) Delta Temperature 0.17 0.34 -0.23 0.10% Failed close-out audit and CARB (°C) audit, invalidated back to previous audit; sensors replaced 12/21/16 Net Radiation 81.45 636.22 -166 91.4% (Wm-2) Precipitation NA NA NA NA Not measured at this station (mm) Relative Humidity 39 100.6 4 96.2% Sensor replaced 12/21/16 (2m, %) Solar Radiation NA NA NA NA Installed Dec 2017 (Wm-2) Sonic 2D WD 360 0 76.6% Removed 12/21/16 (10m, deg) Sonic 2D WS 2.25 11.83 0 76.6% Removed 12/21/16 (10m, ms-1) Wind Direction 181.5 95.3 0.10% Installed 12/21/16 (10m, deg) Wind Direction SD 39.4 96.6 13.35 0.10% Installed 12/21/16 (10m, deg) Wind speed 2.21 11.8 0 71.5% Sensor replaced 12/21/16 (10m, ms-1) Wind speed 0.71 4.8 0.06 71.4% Removed 12/21/16 (1m, ms-1) Wind speed 1.07 6.7 0.07 71.4% Removed 12/21/16 (2m, ms-1)
5 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW
TABLE 5. SONNY BONO METEOROLOGICAL DATA STATISTICS FOR 2016
Parameter Mean Max Min % Recovery Notes
Ambient Temperature 23.16 46.43 0.55 95.8% Sensor replaced 10/3/16 (2m, °C) Barometric Pressure 1.01 1.02 1.00 19.9% Installed 10/3/16 (2m, atm) Delta Temperature 1.02 8.04 -1.86 23.9% Failed close-out audit, invalidated (°C) back to previous audit; sensors replaced 10/3/16 Net Radiation NA NA NA 0.0% Failed CARB audit and was not able (Wm-2) to be recalibrated Precipitation 0.02 5.6 0 19.9% Installed 10/3/16 (mm) Relative Humidity 49.88 100 4.96 95.8% Sensor replaced 10/3/16 (2m, %) Solar Radiation NA NA NA NA Installed Dec 2017 (Wm-2) Sonic 2D WD 359.9 0.4 41.7% Removed 10/3/16 (10m, deg) Sonic 2D WS 4.06 17.94 0.45 41.7% Removed 10/3/16 (10m, ms-1) Wind Direction 359.7 0.04 38.5% Replaced 10/3/16 (this station (10m, deg) already had a vane-style anemometer); Failed close-out audit, invalidated back to previous audit Wind Direction SD 22.5 98.9 2.9 23.9% Installed 10/3/16 (data not (10m, deg) captured with previous vane) Wind speed 3.52 19.09 0.1 95.8% Sensor replaced 10/3/16 (10m, ms-1) Wind speed 1.61 7.03 0 72.0% Removed 10/3/16 (1m, ms-1) Wind speed 1.99 8.93 0 72.0% Removed 10/3/16 (2m, ms-1)
6 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW
TABLE 6. TORRES MARTINEZ METEOROLOGICAL DATA STATISTICS FOR 2016
Parameter Mean Max Min % Recovery Notes
Ambient Temperature 23.38 49.2 -2.7 94.8% Sensor replaced 12/20/16 (2m, °C) Barometric Pressure NA NA NA NA Not measured at this station (2m, atm) Delta Temperature 0.1 0.61 -0.62 0.13% Failed close-out audit and CARB (°C) audit, invalidated back to previous audit; sensors replaced 12/20/16 Net Radiation NA NA NA 0.0% Failed CARB audit and was not able (Wm-2) to be recalibrated Precipitation 0.19 1.1 0 0.13% Installed 12/20/16 (mm) Relative Humidity 43.37 103.2 6.2 90.4% Sensor replaced 12/20/16 (2m, %) Solar Radiation NA NA NA NA Installed Dec 2017 (Wm-2) Sonic 2D WD 359.86 0.1 66.8% Removed 12/20/16 (10m, deg) Sonic 2D WS 2.82 11 0.03 66.8% Removed 12/20/16 (10m, ms-1) Wind Direction 359.7 105.8 0.13% Installed 12/20/16 (10m, deg) Wind Direction SD 37.3 79.47 13.82 0.13% Installed 12/20/16 (10m, deg) Wind speed 2.76 10.9 0.2 75.5% Sensor replaced 12/20/16 (10m, ms-1) Wind speed 1.14 7.1 0.2 24.4% Removed 12/20/16 (1m, ms-1) Wind speed 1.56 7.27 0.11 56.4% Removed 12/20/16 (2m, ms-1)
7 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW 1.2 2016 PARTICULATE DATA HIGHLIGHTS Data recovery was fairly consistent across the six stations over the course of the year. The one notable exception was Salton Sea Park, which had ongoing issues related to the internal logic boards and the cap heater; this was not fully resolved until 2017 Q1. In addition, PM10 at standard temperature and pressure (PM10 STP) data was not calculable until the station transition, since pressure data recorded by the TEOMs were not being reported correctly from the old data loggers.
TABLE 7. BOMBAY BEACH PARTICULATE DATA STATISTICS FOR 2016
Parameter (µgm-3) Mean Max Min % Recovery Notes
PM Coarse 28 6,624 -22.2 88.3%
PM10 local 32.76 7,208 -28.1 88.3%
PM10 STP 17.33 700.3 -13.06 35.7%
PM2.5 4.77 583.1 -29.6 88.3%
TABLE 8. NAVAL TEST BASE PARTICULATE DATA STATISTICS FOR 2016
Parameter (µgm-3) Mean Max Min % Recovery Notes
PM Coarse 79.09 23,191 -42.2 86.0%
PM10 local 90.68 24,535 -39.75 86.0%
PM10 STP 48.62 7,858 -32.2 34.2%
PM2.5 11.53 1,343 -15.24 86.0%
TABLE 9. SALTON CITY PARTICULATE DATA STATISTICS FOR 2016
Parameter (µgm-3) Mean Max Min % Recovery Notes
PM Coarse 100.62 15,578 -61.4 74.3%
PM10 local 114.4 16,606 -59.8 74.3%
PM10 STP 44.59 4,109 -21.91 30.8%
PM2.5 13.72 1,206 -25.5 74.3%
8 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW
TABLE 10. SALTON SEA PARK PARTICULATE DATA STATISTICS FOR 2016
Parameter (µgm-3) Mean Max Min % Recovery Notes
PM Coarse 11.95 140.87 -6.33 1.67%
PM10 local 16.35 167.34 -7.29 1.67%
PM10 STP 10.47 23.16 -5.59 0.11%
PM2.5 4.4 29.13 -9.93 1.67%
TABLE 11. SONNY BONO PARTICULATE DATA STATISTICS FOR 2016
Parameter (µgm-3) Mean Max Min % Recovery Notes
PM Coarse 43.68 7,999 -51.3 78.4% Sudden high loading during a dust storm caused the TEOM to return a “max-out” value of 7999; the same event also caused recovery oscillations in the mass transducer leading to high negative values. Due to the circumstances the data were not invalidated.
PM10 local 50.83 7,999 -58.7 78.4% See note above
PM10 STP 40.19 3,047 -15.31 22.3%
PM2.5 7.35 1,380 -24.7 78.4%
TABLE 12. TORRES MARTINEZ PARTICULATE DATA STATISTICS FOR 2016
Parameter (µgm-3) Mean Max Min % Recovery Notes
PM Coarse 24.61 2,878 -19.1 92.2%
PM10 local 30.04 3,104 -13.66 92.2%
PM10 STP 22.26 1,233 -13.82 37.0%
PM2.5 5.42 625.53 -10.8 92.2%
1.3 2016 ERRATA AND NOTES The following errata and notes are pertinent to the 2016 datasets:
• There are relative humidity measurements at 2m during 2016 Q1 and Q3 at Bombay Beach, 2016 Q1 and Q2 at Salton Sea Park, and 2016 Q1, Q2, and Q3 at Torres Martinez which are greater than 100%. According to the QAPP, “the humidity range of the humidity sensor is from 0-100 percent with an accuracy of ±0.8 percent when the temperature is at 23°C.” However,
9 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW the instruments do drift as they age (up to 1% per year). In addition, the EPA audit tolerance is 8%, so a drift below that will not be flagged for recalibration. After October 2016, the logger program forced the maximum to 100%, but the decision was made to leave values greater than 100% for older data based on the above. • The wind direction measurements at 10m during 2016 Q1 at Sonny Bono do not span full range of 0-360 degrees. Additionally, there is a large gap data in collection (i.e. 2016 Q2 and 2016 Q3 data is missing). Given the full sonic invalidation at Sonny Bono for this time period, and given no other information available to us to invalidate the 10m wind direction (vane) data, the data for the first part of 2016 was retained. There isn’t a full compass span for that limited time frame, approximately ~45 days, but the CARB audit on 2/24/2016 included the vane and indicated passing for all checks so it was presumed that the period 1/1/2016 to 2/24/2016 was valid. Wind direction failed the close-out audit on 10/3/2016 and so data were invalidated back to the last “known good” point, which was the CARB audit on 2/24/2016. • When comparing the wind speed measurements at 1m and 2m, some instruments have a different lower cutoff value to reflect calm or still conditions. The minimum threshold was set inconsistently in the Ecotech system. The minimums for the older data were not standardized, but after October 2016 the logger program is consistent for 10m WS at all sites (1m and 2m WS were removed). • For periods where data are missing, the QA/QC Level is denoted “No QAQC.” This is because, strictly speaking, no QA was done for these periods since data were unrecoverable. However, the data should still be considered “final.” • The vacuum pressure during 2016 Q1 and Q2 at Bombay Beach, Sonny Bono, Salton City, and Torres Martinez appear to be anomalous (i.e. inconsistent with other data). It is believed this is due to a rounding issue with data coming from the Ecotech system. • The PM2.5 mass transducer oscillator noise is essentially zero for most of 2016 Q1 and Q2 at Bombay Beach, Sonny Bono, and Torres Martinez. The data appears distinctly different than other data. This is most likely due to incorrect reporting by the Ecotech system and it wasn’t prioritized for individual flagging during this time period in the review, but it was used as a diagnostic for validating TEOM data if there was reason to believe that the concentration data were suspect. • The vacuum bypass flows 2016 Q1 at Sonny Bono appear to be anomalous (i.e. inconsistent with other data). These are from a data mapping problem from the Ecotech files: 144 is the status code and not the actual data. Similar to the oscillator noise, this parameter was not scrutinized individually as part of validating the TEOM concentrations for the older data and was left "as-is."
10 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW 2 2017 DATA SUMMARY 2.1 2017 METEOROLOGICAL DATA HIGHLIGHTS Basic summary statistics are presented below for each of the six meteorological stations that are part of the SSAQMN. There were no new major issues in 2017 other than the continuing problems with the net radiation sensors at some sites; these were replaced with solar radiation sensors at all six stations in December 2017.
TABLE 13. BOMBAY BEACH METEOROLOGICAL DATA STATISTICS FOR 2017
Parameter Mean Max Min % Recovery Notes
Ambient Temperature 24.47 46.1 3.76 99.0% (2m, °C) Barometric Pressure NA NA NA NA Not measured at this station (2m, atm) Delta Temperature 0.5 7.02 -1.53 92.5% (°C) Net Radiation 72.16 497.8 -102.9 87.1% (Wm-2) Precipitation NA NA NA NA Not measured at this station (mm) Relative Humidity 40.31 96.9 2.81 99.0% (2m, %) Solar Radiation 137.91 662.8 0 8.4% Installed Dec. 2017 (Wm-2) Sonic 2D WD NA NA NA NA Removed 10/4/16 (10m, deg) Sonic 2D WS NA NA NA NA Removed 10/4/16 (10m, ms-1) Wind Direction 360 0.02 99.0% (10m, deg) Wind Direction SD 21.57 100.8 0.01 99.0% (10m, deg) Wind speed 3.21 18.73 0.01 99.0% (10m, ms-1) Wind speed NA NA NA NA Removed 10/4/16 (1m, ms-1) Wind speed NA NA NA NA Removed 10/4/16 (2m, ms-1)
11 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW
TABLE 14. NAVAL TEST BASE METEOROLOGICAL DATA STATISTICS FOR 2017
Parameter Mean Max Min % Recovery Notes
Ambient Temperature 25.33 48.31 2.35 99.2% (2m, °C) Barometric Pressure NA NA NA NA Not measured at this station (2m, atm) Delta Temperature 0.59 7.19 -1.4 94.1% (°C) Net Radiation 0.0% Failed 2016 CARB audit and was (Wm-2) not able to be recalibrated Precipitation NA NA NA NA Not measured at this station (mm) Relative Humidity 33.28 99.5 2.88 99.2% (2m, %) Solar Radiation 129.29 637.7 0 13.0% Installed Dec. 2017 (Wm-2) Sonic 2D WD NA NA NA NA Removed 10/4/16 (10m, deg) Sonic 2D WS NA NA NA NA Removed 10/4/16 (10m, ms-1) Wind Direction 360 0 99.2% (10m, deg) Wind Direction SD 22.15 98.6 1.03 99.2% (10m, deg) Wind speed 4.07 14.89 0.15 29.8% Failed a routine audit and data (10m, ms-1) were back-invalidated to the previous audit. Wind speed NA NA NA NA Removed 10/4/16 (1m, ms-1) Wind speed NA NA NA NA Removed 10/4/16 (2m, ms-1)
12 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW
TABLE 15. SALTON CITY METEOROLOGICAL DATA STATISTICS FOR 2017
Parameter Mean Max Min % Recovery Notes
Ambient Temperature 24.66 47.18 2 99.0% (2m, °C) Barometric Pressure NA NA NA NA Not measured at this station (2m, atm) Delta Temperature 0.57 7.13 -1.88 99.0% (°C) Net Radiation 72.46 468.9 -95.3 85.0% (Wm-2) Precipitation NA NA NA NA Not measured at this station (mm) Relative Humidity 36.82 97.6 2.52 99.0% (2m, %) Solar Radiation 139.62 726.1 0 12.9% Installed Dec. 2017 (Wm-2) Sonic 2D WD NA NA NA NA Removed 10/4/16 (10m, deg) Sonic 2D WS NA NA NA NA Removed 10/4/16 (10m, ms-1) Wind Direction 360 0 99.0% (10m, deg) Wind Direction SD 20.72 99.2 0 99.0% (10m, deg) Wind speed 3.35 18.22 0 99.0% (10m, ms-1) Wind speed NA NA NA NA Removed 10/4/16 (1m, ms-1) Wind speed NA NA NA NA Removed 10/4/16 (2m, ms-1)
13 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW
TABLE 16. SALTON SEA PARK METEOROLOGICAL DATA STATISTICS FOR 2017
Parameter Mean Max Min % Recovery Notes
Ambient Temperature 24.28 44.76 4.34 99.2% (2m, °C) Barometric Pressure NA NA NA NA Not measured at this station (2m, atm) Delta Temperature 0.81 7.38 -2.1 99.2% (°C) Net Radiation 88.63 601.9 -101.1 93.6% (Wm-2) Precipitation NA NA NA NA Not measured at this station (mm) Relative Humidity 38.95 97.8 3.2 99.2% (2m, %) Solar Radiation 132.02 613.9 0 4.75% Installed Dec. 2017 (Wm-2) Sonic 2D WD NA NA NA NA Removed 10/4/16 (10m, deg) Sonic 2D WS NA NA NA NA Removed 10/4/16 (10m, ms-1) Wind Direction 359.9 0.01 99.2% (10m, deg) Wind Direction SD 26.41 101.1 3.03 99.2% (10m, deg) Wind speed 2.25 10.29 0.09 99.2% (10m, ms-1) Wind speed NA NA NA NA Removed 10/4/16 (1m, ms-1) Wind speed NA NA NA NA Removed 10/4/16 (2m, ms-1)
14 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW
TABLE 17. SONNY BONO METEOROLOGICAL DATA STATISTICS FOR 2017
Parameter Mean Max Min % Recovery Notes
Ambient Temperature 23.86 48.47 0.68 96.7% (2m, °C) Barometric Pressure 1.01 1.03 0.99 93.2% (2m, atm) Delta Temperature 0.53 7.79 -2.45 93.9% (°C) Net Radiation 0.0% Failed 2016 CARB audit and was (Wm-2) not able to be recalibrated Precipitation 0 2.3 0 96.7% (mm) Relative Humidity 49.73 100 5.93 96.7% (2m, %) Solar Radiation 133 628.3 0 8.87% Installed Dec. 2017 (Wm-2) Sonic 2D WD NA NA NA NA Removed 10/4/16 (10m, deg) Sonic 2D WS NA NA NA NA Removed 10/4/16 (10m, ms-1) Wind Direction 360 0.03 96.7% (10m, deg) Wind Direction SD 22.31 100 2.51 96.7% (10m, deg) Wind speed 3.17 15.94 0.13 96.7% (10m, ms-1) Wind speed NA NA NA NA Removed 10/4/16 (1m, ms-1) Wind speed NA NA NA NA Removed 10/4/16 (2m, ms-1)
15 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW
TABLE 18. TORRES MARTINEZ METEOROLOGICAL DATA STATISTICS FOR 2017
Parameter Mean Max Min % Recovery Notes
Ambient Temperature 20.23 48.29 -0.51 73.5% (2m, °C) Barometric Pressure NA NA NA NA Not measured at this station (2m, atm) Delta Temperature 0.84 8.83 -1.58 98.4% (°C) Net Radiation 0.0% Failed 2016 CARB audit and was (Wm-2) not able to be recalibrated Precipitation 0.01 10.4 0 98.4% (mm) Relative Humidity 43.22 97.5 4.06 91.0% (2m, %) Solar Radiation 129.22 606.2 0 4.7% Installed Dec. 2017 (Wm-2) Sonic 2D WD NA NA NA NA Removed 10/4/16 (10m, deg) Sonic 2D WS NA NA NA NA Removed 10/4/16 (10m, ms-1) Wind Direction 360 0.01 98.4% (10m, deg) Wind Direction SD 23.23 100.3 3.16 98.4% (10m, deg) Wind speed 2.87 11.23 0.12 98.4% (10m, ms-1) Wind speed NA NA NA NA Removed 10/4/16 (1m, ms-1) Wind speed NA NA NA NA Removed 10/4/16 (2m, ms-1)
16 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW 2.2 2017 PARTICULATE DATA HIGHLIGHTS Many of the TEOMs experienced major repairs in 2017 due to the age of the instruments, which depressed data recovery. Notably the logic board and amplifier boards at Sonny Bono were replaced, which led to a back-invalidation of most of 2017 Q1, and Torres Martinez was offline for all of
September 2017 while waiting for a part. In addition, PM10 STP calculations resumed and incorporated the use of the stand-alone pressure sensor at the Sonny Bono station in the event that the onboard TEOM pressure sensor failed.
TABLE 19. BOMBAY BEACH PARTICULATE DATA STATISTICS FOR 2017
Parameter (µgm-3) Mean Max Min % Recovery Notes
PM Coarse 21.23 3,934 -14.53 95.4%
PM10 local 25.9 4,055 -14.21 95.4%
PM10 STP 25.85 3,957 -14.6 95.4%
PM2.5 4.67 135.6 -12.61 95.4%
TABLE 20. NAVAL TEST BASE PARTICULATE DATA STATISTICS FOR 2017
Parameter (µgm-3) Mean Max Min % Recovery Notes
PM Coarse 37.46 5,562 -16.34 94.2%
PM10 local 44.5 5,922 -14.75 94.2%
PM10 STP 44.84 5,951 -14.67 94.2%
PM2.5 7.06 360 -14.61 94.2%
TABLE 21. SALTON CITY PARTICULATE DATA STATISTICS FOR 2017
Parameter (µgm-3) Mean Max Min % Recovery Notes
PM Coarse 56.8 7,999 -16.66 94.0% Sudden high loading during a dust storm caused the TEOM to return a “max-out” value of 7999. Due to the circumstances the data were not invalidated.
PM10 local 65.38 7,999 -14.72 94.0% See note above
PM10 STP 65.41 7,963 -14.67 94.0%
PM2.5 8.77 572.1 -13.48 94.0%
17 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW
TABLE 22. SALTON SEA PARK PARTICULATE DATA STATISTICS FOR 2017
Parameter (µgm-3) Mean Max Min % Recovery Notes
PM Coarse 18.1 990 -14.57 83.2%
PM10 local 23.35 1,099 -14.74 83.2%
PM10 STP 23.24 1,096 -15.04 83.2%
PM2.5 5.25 157.9 -13.52 83.2%
TABLE 23. SONNY BONO PARTICULATE DATA STATISTICS FOR 2017
Parameter (µgm-3) Mean Max Min % Recovery Notes
PM Coarse 36.13 1,119 -19.07 70.3%
PM10 local 43.87 1,164 -14.89 70.3%
PM10 STP 43.62 1,165 -14.99 70.3%
PM2.5 7.74 247.2 -14.85 70.3%
TABLE 24. TORRES MARTINEZ PARTICULATE DATA STATISTICS FOR 2017
Parameter (µgm-3) Mean Max Min % Recovery Notes
PM Coarse 29.06 6,611 -14.48 76.2%
PM10 local 35.1 6,803 -14.6 76.2%
PM10 STP 35.03 6,730 -14.63 76.2%
PM2.5 6.05 191.7 -12.14 76.2%
2.3 2017 ERRATA AND NOTES The following errata and notes are pertinent to the 2017 datasets:
• The PM2.5 mass accumulations are negative during 2017 Q3 at Salton City and 2017 Q1, Q2 and Q3 at Salton Sea Park. This correlates with brief (1-hour) periods where the TEOM calculates a negative concentration due to an instability in the instrument. This does not invalidate the concentration data (other than the brief period of extreme negative concentration) since the concentration data is a different calculation stream than the mass accumulation. High relative humidity can contribute to this and an example of this occurred at Salton City in 2017. If the incoming PM is hydrated, sticks to the filter, and then dries after a brief period, it will destabilize the calculations for that hour, and since the calculation is simply a subtraction from the previous
18 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW hour, the instrument can end up with negative mass. According to Thermo Scientific, the manufacturer of the TEOMs, brief power interruptions can also affect this calculation. • The vacuum bypass flows during 2017 Q2 at Naval Test Base appear to be anomalous (i.e. inconsistent with other data). This period of "jitter" in 2017 corresponds to a period with a number of errors and site visits, including an audit where the flows were adjusted, but were still within tolerance (12 l/min +/- 5%).
19 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW 3 SUMMARY STATISTICS 2010–2017 The annual percent data recovery by station and parameter are summarized in the table below.
TABLE 25. RUNNING SUMMARY OF DATA RECOVERY SINCE THE NETWORK WAS INSTALLED Bombay Beach (BB)
Year WS WS WS Diff AT RH Net WD Precip BP PM10 1M 2M 10M Temp 2m 2m Radβ 10mχ 2m Local 2010δ 99.4% 99.4% 99.0% 99.3% 99.4% 99.4% 98.9% 99.4% NA NA 75.0% 2011 99.4% 88.6% 91.9% 95.8% 99.4% 96.5% 95.8% 99.4% NA NA 43.5% 2012 99.6% 59.1% 99.6% 35.3% 99.6% 99.6% 98.2% 92.9% NA NA 64.1% 2013 35.4% 98.9% 95.7% 36.5% 98.9% 98.9% 98.9% 71.4% NA NA 91.5% 2014 66.7% 98.2% 96.8% 98.3% 98.2% 98.2% 98.3% 63.2% NA NA 75.8% 2015α 98.8% 98.8% 98.8% 98.1% 98.8% 97.7% 98.1% 59.0% NA NA 70.4% 2016 70.2% 70.2% 92.2% 22.0% 36.8% 92.1% 11.0% 89.6% NA NA 88.3% 2017 NA NA 99.0% 92.5% 99.0% 99.0% 87.1% 99.0% NA NA 95.4% Overall 71.2% 76.7% 96.6% 72.2% 91.3% 97.7% 85.8% 84.2% NA NA 75.5% Naval Test Base (NTB)
Year WS WS WS Diff AT RH Net WD Precip BP PM10 1M 2M 10M Temp 2m 2m Radβ 10mχ 2m Local 2010δ 99.1% 99.1% 98.9% 99.1% 99.1% 99.1% 98.0% 99.1% NA NA 88.0% 2011 99.2% 99.2% 80.9% 99.2% 99.2% 99.2% 99.2% 97.3% NA NA 91.9% 2012 86.2% 91.6% 85.2% 90.5% 99.3% 99.3% 90.2% 97.7% NA NA 82.1% 2013 96.9% 96.2% 89.2% 95.9% 96.9% 96.9% 93.7% 62.5% NA NA 63.1% 2014 90.5% 92.5% 93.8% 97.2% 97.2% 97.2% 97.2% 81.7% NA NA 85.5% 2015α 73.0% 95.3% 95.3% 94.5% 95.3% 95.3% 95.2% 95.4% NA NA 92.7% 2016 60.6% 68.6% 89.0% 17.6% 31.3% 88.9% 0.0% 67.1% NA NA 86.0% 2017 NA NA 29.8% 94.1% 99.2% 99.2% 0.0% 99.2% NA NA 93.8% Overall 75.7% 80.3% 82.8% 86.0% 89.7% 96.9% 71.7% 87.5% NA NA 85.4% Salton City (SC)
Year WS WS WS Diff AT RH Net WD Precip BP PM10 1M 2M 10M Temp 2m 2m Radβ 10mχ 2m Local 2010δ 91.6% 93.4% 92.2% 93.4% 93.4% 93.4% 90.7% 93.4% NA NA 80.9% 2011 55.8% 69.6% 81.4% 81.9% 81.9% 81.9% 74.4% 82.5% NA NA 73.3% 2012 23.5% 40.8% 45.7% 46.8% 46.8% 46.8% 31.9% 91.1% NA NA 31.9% 2013 38.8% 79.9% 84.4% 85.7% 86.6% 86.6% 22.8% 57.3% NA NA 47.5% 2014 45.1% 72.8% 80.5% 80.5% 80.5% 72.0% 80.5% 71.8% NA NA 52.2% 2015α 73.4% 91.8% 91.8% 91.7% 91.8% 91.7% 91.6% 42.2% NA NA 86.1% 2016 25.8% 25.8% 82.3% 21.7% 82.3% 82.3% 82.3% 61.3% NA NA 74.3% 2017 NA NA 99.0% 99.0% 99.0% 99.0% 85.0% 99.0% NA NA 93.0% Overall 44.3% 59.3% 82.2% 75.1% 82.8% 81.7% 69.9% 74.8% NA NA 67.4%
20 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW
Salton Sea Park (SSP)
Year WS WS WS Diff AT RH Net WD Precip BP PM10 1M 2M 10M Temp 2m 2m Radβ 10mχ 2m Local 2010δ 97.8% 97.8% 97.8% 97.8% 97.8% 97.8% 95.0% 97.8% NA NA 83.0% 2011 92.1% 92.1% 80.4% 92.1% 92.1% 92.1% 85.3% 68.5% NA NA 87.2% 2012 99.6% 99.6% 99.6% 76.4% 99.6% 99.6% 71.5% 99.0% NA NA 64.4% 2013 94.2% 94.2% 94.2% 94.1% 94.2% 94.2% 85.1% 85.1% NA NA 86.4% 2014 85.2% 85.2% 85.2% 99.2% 92.0% 99.2% 91.1% 86.2% NA NA 83.6% 2015α 92.4% 95.6% 95.6% 95.7% 95.6% 95.6% 95.7% 73.9% NA NA 85.8% 2016 71.4% 71.4% 71.5% 0.10% 14.9% 96.2% 91.4% 76.7% NA NA 1.67% 2017 NA NA 99.2% 99.2% 99.2% 99.2% 93.6% 99.2% NA NA 82.6% Overall 79.1% 79.5% 90.4% 81.8% 85.7% 96.7% 88.6% 85.8% NA NA 71.8% Sonny Bono (SB)
Year WS WS WS Diff AT RH Net WD Precip BP PM10 1M 2M 10M Temp 2m 2m Radβ 10mχ 2m Local 2010δ 0.0% 0.0% 0.0% 97.5% 97.8% 97.8% 96.0% 99.2% NA NA 82.4% 2011 0.0% 0.0% 0.0% 98.8% 97.9% 89.7% 98.8% 98.9% NA NA 66.2% 2012 0.0% 0.0% 0.0% 14.9% 53.6% 56.3% 68.5% 88.8% NA NA 87.8% 2013 0.0% 0.0% 0.0% 66.9% 84.4% 89.6% 86.9% 92.0% NA NA 87.5% 2014 0.0% 0.0% 0.0% 65.4% 65.8% 65.8% 65.7% 76.6% NA NA 46.4% 2015α 0.0% 0.0% 0.0% 24.3% 24.3% 24.3% 24.3% 25.9% NA NA 29.4% 2016 72.0% 72.0% 95.8% 23.9% 95.8% 95.8% 0.0% 41.7% NA NA 78.4% 2017 NA NA 96.7% 93.9% 96.7% 96.7% 0.0% 96.7% 96.7% 93.2% 69.6% Overall 9.0% 9.0% 24.1% 60.7% 77.0% 77.0% 55.0% 77.5% 96.7% 93.2% 68.5% Torres-Martinez (TM)
Year WS WS WS Diff AT RH Net WD Precip BP PM10 1M 2M 10M Temp 2m 2m Radβ 10mχ 2m Local 2010δ 71.9% 71.9% 71.8% 71.7% 71.8% 71.8% 40.6% 71.9% NA NA 58.2% 2011 91.0% 91.0% 91.0% 91.0% 91.0% 91.0% 82.7% 91.0% NA NA 4.7% 2012 89.3% 89.3% 89.3% 84.0% 89.3% 89.3% 43.7% 88.6% NA NA 20.4% 2013 94.1% 51.7% 94.1% 91.8% 94.1% 94.1% 1.3% 63.9% NA NA 49.7% 2014 82.7% 47.6% 97.4% 97.4% 97.4% 97.4% 84.8% 77.9% NA NA 90.0% 2015α 97.9% 97.9% 97.9% 96.9% 91.7% 91.7% 96.9% 46.1% NA NA 88.2% 2016 24.4% 56.4% 75.5% 0.13% 94.8% 90.4% 0.0% 66.8% NA NA 92.2% 2017 NA NA 98.4% 98.4% 73.5% 91.0% 0.0% 98.4% 98.4% NA 76.9% Overall 68.9% 63.2% 89.4% 78.9% 88.0% 89.6% 43.8% 75.6% 98.4% NA 60.0% α 2015 did not undergo the same level of QA/QC as other years. β Net radiation sensors were replaced with solar radiation sensors in December 2017. See Section 2.1. χ Wind direction was measured by sonic anemometers until late 2016 when they were replaced with vanes. Recovery in 2016 combines sonic anemometer values with vane values. See Section 1.1. δ Data in 2010 was collected from February 1 to December 31.
21 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW 4 GENERAL DATA REVIEW QA/QC METHODS After monitored data is gathered and imported into the Salton Sea Air Quality Monitoring Program Data Portal (data portal), the Team follows a multi-level QA/QC process to meet data quality objectives. The Team utilizes automated data review procedures and the expertise of numerous technical staff to review and quality check data from the monitoring network.
The Salton Sea Air Quality Team’s data review staff include: the field crews/monitoring personnel responsible for proper operation of the meteorological and air quality stations, other technical personnel with direct knowledge of the proper operation of the monitoring equipment and data gathered, and staff and senior-level database programmers and air quality scientists.
The Team’s multi-stage data review QA/QC process is described below:
Data Review, Level 0: After data is transmitted to the Team and loaded into the data portal, it is assigned as Level 0 data. At this stage, data are available for simple plotting and tabulations to ensure that the data from the instruments are being recorded and properly transmitted to the Team. Any issues with the monitoring equipment can be quickly identified by members of the Team.
Data Review, Level 1: At this review stage, the Salton Sea Air Quality Team’s technical staff review the data in the data portal to ensure that there are no missing data periods. If there are missing data periods, the staff consult with the field crews and database programmers to determine if the data is available (or can be directly downloaded from the station dataloggers) and can be uploaded to the data portal. If the Team cannot fill in the missing data periods, the data is flagged as missing and unrecoverable and is invalidated.
For the Level 1 reviews, the data is reviewed for completeness and general reasonableness and proposed invalidation flags are placed in the data portal to indicate potential data invalidations. In addition, the data reviewer ensures that documentation from site visits to the monitoring stations, including any audit/calibration activities, are entered in the data portal and are flagged so that data recorded during these activities can be invalidated.
Data Review, Level 2: At this review stage, senior-level scientists review the Level 1 data and accept/reject the proposed data invalidations from the Level 1 review. For the Level 2 review, the Level 1 data is reviewed to eliminate outlier/clearly anomalous data and to ensure that the data collected is of good quality.
First, the Level 1 data is output from the data portal into Microsoft Excel or other programs so that the data files can be visually inspected line-by-line. The Level 1 datafiles are reviewed to determine that the proposed data invalidation flags are correctly aligned with the data to be invalidated. Also, additional invalidation flags, not recorded during the Level 1 review, may be added as needed.
22 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW Second, any audit and calibration information is considered. If a particular instrument failed an audit or calibration during the data review period, data from the failed instrument/parameter is typically back invalidated to the most recent, previous passing audit or calibration.
After the review of the Level 1 data files, the Level 2 reviewer generates numerous graphical plots to visually inspect the data. The graphical plots are intended to display any clearly erroneous data that may have not have been flagged during the Level 1 review. In addition, the Level 2 reviewer relies on his/her knowledge of atmospheric and air pollution processes to determine if data may be suspect or erroneous and in need of invalidation. The data from the group of monitoring stations are reviewed collectively and on an individual station basis.
Meteorological Data: The following information is reviewed/considered for the data from the meteorological monitoring stations:
Wind Speed, Wind Direction, and Standard Deviation of Horizontal Wind Direction (sigma theta):
• Maximum and minimum daily wind speed values - minimum wind speeds should be greater than or equal to 0 and maximum winds should be explained by meteorological conditions.
• Maximum and minimum daily wind direction values - minimum wind directions should greater than or equal to 0 degrees and maximum wind directions should be no more than 360 degrees.
• The data portal automatically flags the wind speed and direction data and outputs a warning review flag if data values are constant for more at least three hours. In these cases, the reviewer is to determine if the data in valid/invalid. Some examples where data might be considered invalid might consist of an unchanging wind data due to instrument damage (i.e., a wind vane being damaged, a frozen anemometer, bearings issues with the equipment causing it not to move, etc.).
• Standard deviation of horizontal wind direction values - minimum sigma theta values should be greater than or equal to 0 degrees and it is the Team’s experience that maximum sigma theta values are generally around about 90 - 100 degrees; however, sigma theta values greater than 100 degrees can and do occur and can be explained by wind vane meander under very low wind speed conditions. If there are some higher values or clear outliers, the sigma theta data are reviewed further in the context of the associated meteorological conditions to make a valid/invalid determination.
Ambient Temperature and Relative Humidity:
• Maximum and minimum daily data values are reviewed to ensure the data is reasonable given expected climatological normals.
• Diurnal patterns are reviewed to determine if the minimum temperature occurs in morning, and the maximum temperature occurs in the afternoon; if this pattern is not present, further analysis of recent cold front passages, stormy conditions/clouds, etc. are reviewed to explain the data further.
23 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW Delta Temperature (10-meter temperature minus 2-meter temperature):
• Diurnal patterns are reviewed - typical pattern is negative values (or weak positive values) in the daytime when the surface is generally warmer than the air above it (i.e., 2-meter temp. > 10-meter temp.) and positive values in the nighttime when the surface is colder than the air above it (i.e., 10-meter temp. > 2-meter temp. – nighttime atmospheric inversion).
Net Radiation:
• Maximum and minimum daily data values are reviewed to ensure the data is reasonable given expected climatological normals.
• Diurnal patterns are reviewed to determine if the minimum net radiation value occurs overnight (and are negative), and the maximum net radiation value occurs during the daytime; if this pattern is not present, further analysis of stormy conditions/clouds, etc. are reviewed to explain the data further.
Meteorological Data (Sonic Anemometer): Prior to the Team’s operation of the meteorological and air quality stations starting in late 2016, the Salton Sea meteorological network also consisted of a sonic anemometer system at each of the six monitoring stations. After the Team took over the stations, the sonic anemometer systems were discontinued.
The Team has reviewed the previously-collected sonic anemometer data from January 1, 2016 until late 2016 in the event the data can be used for future modeling and analysis purposes. It was important to preserve as much of the sonic data as possible (wind direction in particular). Prior to the Team’s operation of the stations, the meteorological network was not configured to gather wind direction data from a conventional wind vane; with the exception of Sonny Bono, wind direction data was only gathered from the sonic anemometer.
The following is a summary of the Team’s data review of the sonic anemometer data for 2016:
• Data invalidation flagging was made mostly consistent with the SSAQMN (i.e., Data Check Notes) data review logs that were compiled by the network operator from January to September 2016 and this information was supplemented with additional invalidations were needed.
• Consistent with the Data Check Notes, all periods with a Young Error (an error code self- reported by the sonic anemometer) greater than zero were invalidated for all sonic anemometer parameters.
• All azimuth wind directions were invalidated for 2016. Per the Team’s review of this data, the northerly wind directions were missing so that wind directions didn’t span a full 360 degrees, resulting in incomplete wind roses. In lieu of the azimuth wind direction data, future data users could potentially use the available 2-D wind direction data.
• Given data issues at Sonny Bono from January 1 to April 22, 2016, the following parameters were invalidated by the Team: 2-D and 3-D vector average wind directions (wind directions
24 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW didn’t span a full 360 degrees; missing from the north), 2-D and 3-D vector average wind speeds, and 2-D and 3-D sigma theta values.
• Erroneous vertical wind components greater than around 1 meter per second were invalidated. There were many instances in the data sets where the vertical wind component was highly erroneous/anomalous and spiked starting around hour 00:00 and then slowly dropped back to baseline over the course of a few days. The Team cleaned up and invalidated the suspected erroneous data as much as possible, but the Team has little confidence in the data for this parameter and doesn’t plan on using this data in future work.
• A few erroneous sonic anemometer temperature values were invalidated.
Particulate Data: The following general information is reviewed/considered for the data from the air quality (TEOM) monitoring stations:
Particulate concentration data (PM2.5, PM10 local conditions, PM10 STP, PM Coarse):
• Particulate concentrations are reviewed - the higher concentration values are usually explained by high winds. In some cases, local activities or forest fire influences can explain the higher concentrations.
• In general, the PM concentrations are not invalidated unless there is a strong reason to do so. Possible reasons for invalidation might include significant/unreasonable data outliers (both strongly positive or negative), erroneous constant values such as 7999, values where the TEOM is indicating a status or operational mode error and the values are stuck at constant values, etc.
• Based on conversations with Thermo Scientific, the manufacturer of the TEOMs, the Team invalidates hourly PM concentrations less than -15 µg/m3. In some selected instances when there is a very big dust event and very high PM concentration readings occur, as the TEOM stabilizes from the high mass loading, the Level 2-validated, post-dust event concentration data may contain a few PM concentrations less than -15 µg/m3 and the Team might not invalidate these data.
• PM2.5 and PM10 are reviewed concurrently for each station in the network to ensure that, in general, the PM10 concentration data are greater than the PM2.5 concentration data.
Both relative humidity and temperature readings measured from the TEOM are reviewed in a similar fashion to the relative humidity and temperature readings from the meteorological station. Barometric pressure is also reviewed and is straightforward to review since atmospheric pressures should read around 1 atmosphere (atm).
25 SALTON SEA - AMBIENT AIR QUALITY MONITORING NETWORK 2016-2017 DATA REVIEW 5 REFERENCES Environmental Protection Agency. 2000. “Meteorological Monitoring Guidance for Regulatory Modeling Applications”. EPA-454/R-99-005. February 2000. https://www3.epa.gov/scram001/guidance/met/mmgrma.pdf. Accessed August 18, 2017.
—. 2007. “Guidance for Preparing Standard Operating Procedures (SOPs)”. EPA/600/B-07/001. April 2007. http://www.epa.gov/QUALITY/qs-docs/g6-final.pdf. Accessed May 22, 2015.
—. 2013. “Quality Assurance Handbook for Air Pollution Measurement Systems: Volume II Ambient Air Quality Monitoring Program”. EPA-454/B-13-003. May, 2013. https://www3.epa.gov/ttnamti1/files/ambient/pm25/qa/QA-Handbook-Vol-II.pdf. Accessed August 18, 2017.
—. 2016a. "Guidance on the Development of Modeled Emission Rates for Precursors (MERPs) as a Tier 1 Demonstration Tool for Ozone and PM2.5 under the PSD Permitting Program." Memorandum from Richard A. Wayland, EPA Air Quality Assessment Division Director, to EPA Regional Air Division Directors, Regions 1-10, December 2.
—. 2016b. "Guidance on the Use of Models for Assessing the Impacts of Emissions from Single Sources on the Secondarily Formed Pollutants: Ozone and PM2.5." EPA-454/R-16-005. EPA Office of Air Quality Planning and Standards, Air Quality Assessment Division, December.
—. 2017. "Revision to the Guideline on Air Quality Models: Enhancements to the AERMOD Dispersion Modeling System and Incorporation of Approaches to Address Ozone and Fine Particulate Matter." 40 CFR Part 51, Appendix W. Docket ID No. EPA-HQ-OAR-2015-0310, January 17.
26