Skip to content

Latest commit

 

History

History
416 lines (315 loc) · 18.3 KB

readme.md

File metadata and controls

416 lines (315 loc) · 18.3 KB
Raw

Mobile Air Chemistry Laboratory at Sunset Park

Lewiston-Clarkston Valley Formaldehyde Study (2016)

Online: https://bitbucket.org/wsular/2016-hcho-macl-daq

Documentation for the Mobile Air Chemistry Laboratory (MACL) at Sunset Park. The MACL contains gas-phase analyzers and meteorological sensors; measurements include:

  • At ~36ft (11m), by ultrasonic weather station (WXT510; Vaisala)
    • wind direction (avg/min/max)
    • wind speed (avg/min/max)
    • barometric pressure
    • air temperature
    • relative humidity
    • rain (accumluation/duration/intensity)
    • hail (accumulation/duration/intensity)
  • At ~22ft (6.7m), via 1/2" I.D. PFA tubing sampling line by trace gas analyzers
  • At ~20ft (6.1m), by some kind of NOy analyzer (some company) TODO FIXME
    • reactive oxides of nitrogen (NOy)
  • Within the trailer

Additionally, a ceilometer (CL31; Vaisala) deployed on the MACL roof captures cloud height and coverage. Usage and data handling for the ceilometer are not covered here.

The PTR-MS and NOy analyzer are listed for completeness; neither instrument is addressed here.

Acquisition

Data is aggregated into several locations. Where possible, measurements are acquired redundantly to prevent data loss and provide the highest-quality data signals. The higher-quality source is listed in the "primary DAQ" column in the table below.

Measurements Primary DAQ Secondary DAQ
meteorology (WXT510) DAQFactory not applicable
sample mass flow DAQFactory not applicable
trace gases (Teledynes) APICOM DAQFactory
trace gases (LI-840A) LI-840A Software DAQFactory

DAQFactory

A 19" rack-mounted PC running HMI/SCADA software (DAQFactory 5.87a; Azeotech) acquires data via analog voltage signals (for gas analyzers and flow meter) or serial port (for weather station) and stores data to delimited text file. Raw measurements are aggregated on 60sec intervals into mean (average) values prior to storage. A new file is created each day at midnight.

N.B. All measurements are aggregated into mean values within the 1-minute logging interval, including wind direction avg/min/max and relative humidity from the WXT510.

 

Analog voltage signals are inherently susceptible to noise and bias. At low concentrations, the signal may be distorted significantly. To avoid this, it is recommended to use data in files created by APICOM or LI-840A Software in preference to DAQFactory data files.

Here is a screenshot of the computer with DAQFactory running on the 2nd desktop (multiple desktops provided by VirtuaWin).

Screenshot of DAQFactory on desktop 2

APICOM

The Teledyne analyzers are configured to retain data in non-volatile onboard memory:

  • ~7 days of 1-min mean values (concentration(s) + stability indicator(s))
  • ~31 days of 30-min mean values (same params)

Several instances of the manufacturer-provided communications utility (APICOM 5.0.5; Teledyne API) run in the background on the rack PC, periodically retrieving the latest records and appending them to cumulative delimited text files. One instance handles the retrieval for one analyzer.

Here is a screenshot of the APICOM software running on desktop #3:

Screenshot of APICOM on desktop 3

LI-840A Software

A single instance of the manufacturer-provided interface software (LI-840A Software 2.0.0; LI-COR Biosciences) runs in the background on the rack PC, performing continuous acquisition of the serial data stream. The data is recorded without aggregation into a cumulative delimited text file.

In the interest of long-term reliability, do not engage the plotting features of LI-840A Software.

Here is a screenshot of the LI-840A Software running on desktop #4:

Screenshot of LI-840A Software on desktop 4

Data Products

Examples

See this Jupyter notebook (rendered with nbviewer):

DAQFactory Files

  • File name format: Lewiston_RackYYMMDD.txt where YYMMDD is date stamp using two-digit year
  • Record interval: nominal 1-min (every 60sec/not aligned to minute) (anchors unknown)
  • Aggregation: mean 1-min values of real-time measurements
    • naive mean used, including for WD & RH from WXT510
Column name Units Description
TheTime UTC-0800 seconds since midnight Jan 1, 1970 in local time (Pacific Standard Time)
LabjackU3_T degC temperature of DAQ hardware
LI840A_CO2 ppm carbon dioxide mixing ratio
LI840A_H2O ppth water vapor mixing ratio
TAPI_CO ppb carbon monoxide mixing ratio
TAPI_NO ppb nitric oxide mixing ratio
TAPI_NO2 ppb nitrogen dioxide mixing ratio
TAPI_NOx ppb total mono-nitrogen oxides mixing ratio
TAPI_O3 ppb ozone mixing ratio
TAPI_SO2 ppb sulfur dioxide mixing ratio
TSI_flow volts sampling inlet flow (unscaled)
WXT_Dm m/s wind direction average
WXT_Dn m/s wind direction minimum
WXT_Dx m/s wind direction maximum
WXT_Hc hits/cm^2 hail accumulation
WXT_Hd s hail duration
WXT_Hi hits/hr*cm^2 hail intensity
WXT_Hp hits/hr*cm^2 hail peak intensity
WXT_Pa mb air pressure
WXT_Rc mm rain accumulation
WXT_Rd s rain duration
WXT_Ri mm/hr rain intensity
WXT_Rp mm/hr rain peak intensity
WXT_Sm m/s wind speed average
WXT_Sn m/s wind speed minimum
WXT_Sx m/s wind speed maximum
WXT_Ta degC air temperature
WXT_Ua % relative humidity

Example: convert timestamp to Excel serial date (assuming TheTime is column A) (ref):

=(A1/86400)+DATE(1970,1,1)

Example: import using pandas in Python:

import pandas as pd

df = pd.read_csv('/path/to/the/file.txt',
                 header=0, # headers 1st row
                 index_col=0, # timestamps 1st column
                 # specify as seconds to recognize as epoch timestamp.. BUT
                 # it's local time (i.e. no timezone conversion required)
                 # (Section 9.10.3, http://www.azeotech.com/dl/usersguide.pdf)
                 date_parser=lambda x: pd.to_datetime(x, unit='s'),
                 parse_dates=True)

APICOM Files

  • File name format: XX_TTT_YYYYMMDD.csv where
    • XX is the measured gases (CO, NONO2, O3 or SO2)
    • TTT is either of 1min or 30min, representing 1-min or 30-min values
    • YYYYMMDD is the file creation date (not start of data in file)
  • Record interval:
    • either 1-minute or 30-minutes (closed left, label right)
    • in both cases, timestamped at 1-second resolution (not 1-minute) and thus label is of form hh:mm:01
  • Aggregation: mean 1-min values of real-time measurements
CO
Column name Units Description
Time Stamp UTC-0800 timestamp in US date format (mm/dd) (PST)
CONC1-AVG (PPM) ppm mean carbon monoxide mixing ratio
STABIL-AVG (PPM) ppm concentration stability
NONO2
Column name Units Description
Time Stamp UTC-0800 timestamp in US date format (mm/dd) (PST)
NOXCNC1-AVG (PPB) ppb mean total mono-oxides of nitrogen mixing ratio
NOCNC1-AVG (PPB) ppb mean nitric acid mixing ratio
NO2CNC1-AVG (PPB) ppb mean nitrogen dioxide mixing ratio
STABIL-AVG (PPM) ppb* concentration stability

* File header contradicts common sense and user manual -- it's probably ppb.

O3
Column name Units Description
Time Stamp UTC-0800 timestamp in US date format (mm/dd) (PST)
CONC1-AVG (PPM) ppm mean ozone mixing ratio

TODO where is STABIL? was supposed to be included in recorded parameters but is missing? updating now will dump existing data set from memory...

SO2
Column name Units Description
Time Stamp UTC-0800 timestamp in US date format (mm/dd) (PST)
CONC1-AVG (PPM) ppm mean sulfur dioxide mixing ratio
STABIL-AVG (PPM) ppm concentration stability

Example: import using pandas in Python (valid for all TAPI csv files):

import pandas as pd

df = pd.read_csv('/path/to/the/file.txt',
                 header=0, # headers 1st row
                 index_col=0, # timestamps 1st column
                 parse_dates=True, # US format (MM/DD/YYYY)
                 skipinitialspace=True) # spaces and commas

LI-840A Software Files

  • File name format: li840a_YYYYMMDD.txt where YYYYMMDD is the file creation date (also typ. start of data in file)
  • Record interval: 1 second
  • Aggregation: none; represents real-time measurements
Column name Units Description
Date(Y-M-D) UTC-0800 date portion of timestamp (PST)
Time(H:M:S) UTC-0800 time portion of timestamp (PST)
CO2(ppm) ppm carbon dioxide mixing ratio
H2O(ppt) ppth water vapor mixing ratio
H2O(C) degC water vapor dewpoint
Cell_Temperature(C) degC temperature in measurement cell
Cell_Pressure(kPa) kPa pressure in measurement cell

Example: import using pandas in Python:

import pandas as pd

df = pd.read_table('/path/to/the/file.txt', # table for tab-separated
                   header=1, # headers 2nd row
                   index_col=0 # dates 1st col, times 2nd col
                   parse_dates={'timestamp': [0,1]}) # combine cols

Initial Device Setup

Step-by-step instructions are available in the relevant user manuals. See References below.

Weather Station (WXT510)

Little, if any, configuration is required.

  • wired as RS-232 and powered through modified USB-serial adapter
  • used in polled mode

Teledyne analyzers

General config done (i.e. optional, not project-specific):

  • give nice hostnames (co, nox, ...)
  • use DHCP (in conjunction with IP reservations)

Project-specific configuration files for onboard data logging (iDAS feature) for each analyzer are available under /teledyne-idas. Load these config files using APICOM.

Router

Things to do:

  • create DHCP reservations (static IP address assignments)
  • create port forward rules for Remote Desktop Connection
  • disable Wi-Fi

Broadband Modem (Raven X)

Login to the device's web interface (typ. http://192.168.13.31:9191 from the LAN side) and set the following configuration (on factory defaults):

  • LAN > Ethernet
    • Host Public Mode: Ethernet Uses Public IP
  • LAN > Global DNS (optional, recommended)
    • Alternate Primary DNS:
    • Alternate Secondary DNS:
  • Services > ACEManager (optional, strongly recommended for emergency remote administration)
    • OTA ACEmanager Access: SSL Only
  • Services > Telnet/SSH
    • AT Server Mode: SSH
    • [Make SSH Keys] button: Press if SSH is not initalized
  • Services > Time (SNTP)
    • Enable time update: Enabled
    • SNTP Server Address: pick one

Apply and reboot.

Rack PC

The rack PC runs DAQFactory, APICOM, and LI-840A Software for data acquisition purposes -- see above for details. Additionally, the rack PC runs:

  • Network Time Protocol Service, to keep clock tightly synchronized
  • WinSCP, to mirror local data files to a WSU-hosted server

We are omitting all the specific details of setup for the rack PC, sorry.

Instead of using the "keep-up-to-date" feature of WinSCP (which would result in a massive cell phone bill), we run it periodically with Windows Task Scheduler. Here is a (tastefully redacted) screenshot of WinSCP in action:

Screenshot of WinSCP on desktop 1

And here are the contents of the batch file used to launch WinSCP:

option batch on
option confirm off
open macl@lar-field-data
synchronize remote -mirror -criteria=either C:\LCV_HCHO_2016\data /2016_HCHO/data/sunsetpark
exit

References