SOP - Light Absorption Coefficient - Filter-based - Radiance Research Particle/Soot Absorption Photometer (PSAP).

From TROPOS Wiki

Jump to: navigation, search
Standard Operating Procedures for
Particle Soot Absorption Photometer (PSAP)
for the Global Atmospheric Watch (GAW) Programme

Return to Standard Operating Procedures



This document outlines the standard operating procedures for the measurement of aerosol light absorption coefficients using a Particle Soot Absorption Photometer (PSAP, Radiance Research, Inc., Seattle, WA, USA).  There is a need for the standard operating procedures applicable to all WMO/GAW sites. The aerosol light absorption measurement has been recommended by Scientific Advisory Group on Aerosols for long-term measurements in the Global Atmospheric Watch global network, as a subset of essential core variables (GAW Report 153, 2003).  Having a common guideline for the operation and maintenance of PSAP instruments at all GAW sites ensures a similar quality for light absorption measurements and to reduce potential systematic biases in the data due to differences in methodologies, set-up and maintenance.

Figure 1: 1W & 3W PSAP (Radiance Research Inc.) as installed at Alert GAW station.

Principle of PSAP Operation

PSAP is used to measure high time-resolution aerosol light absorption coefficient, σap. This measurement is based on the rate of change of light transmission through a fiber filter, as particles deposit and accumulate onto the filter.  There are two types of PSAP instruments commercially available at Radiance Research Inc.; a single-wavelength PSAP that measures light absorption at the wavelength of 565 nm and a 3-wavelength PSAP (3W PSAP) that measures light absorption at wavelengths of 467, 530 and 660 nm respectively.  Air is drawn through two identical glass fiber filters in series and an aerosol sample is collected on the exposed area (~ 0.5 cm in diameter) of the first filter.   Light is transmitted through the aerosol sample on the first filter (sample filter) and through the clean second filter (reference filter). The light intensities detected are compared between the sample side and the clean reference side.  As aerosols accumulate on the sample filter with time the light transmittance drops and light absorption coefficients are determined by the Beer’s-law:




where Area=exposed sample spot, Vol is volume of air sampled, I0 is filter transmittance in average period i and I is transmittance in period (i+1).  The software within the instrument has an algorithm which corrects for the filter loading non-linearity.  This transfer function f(τ) is incorporated in the software and is based on filter loading for Pallflex filters (Pall Life Sciences, E70-2075W).  The function is defined by Equation 2:




where τ is filter transmission (set to unity in the software for unloaded filter). It should be noted that Equation (2) is applicable only to Pallflex filters.  If other filters are to be used, Equation (2) must be re-evaluated.  At this time, the PSAP manufacturer has only recommended the use of the Pallflex filters.

Inspection and Set-up of PSAP

Initial inspection and checks

A new instrument as received from the company should go through an initial inspection as some electrical components might become loose in transportation.

1). Remove the top panel and ensure all the electrical connections are in place and there are no loose cables.


2). Ensure that the tube connections follow the actual flow schematic. In particular, make sure that the flow path permits air flow from the outlet (bottom) of the sample cell to the inlet (top) of the reference cell.  The flow through the PSAP instrument should be in the order as shown below.


 Sample inlet→ Sample cell→ Reference cell→ Flow meter

 Needle valve→ Outlet→ Pump

Figure 2: This picture shows the electronics and tubing connections inside the PSAP. Red arrows indicate the direction of the flow in a correct orientation.

3). Conduct a leak test.  Turn the instrument on.  Leak test must be performed as follows before putting the instrument into operation:


Put a filter in the filter-holder.  Cap off the sample inlet, connect and turn on the pump on the vacuum side.  The flow on the PSAP display (or in data captured via the serial port) should be close to zero if there are no leaks in the system.  This can also be done by a handheld leak checker, if available, with a vacuum gauge.  The leak rate should be low and should not exceed a few millibars per minute.


4). Overnight zero test must be performed on a new PSAP by connecting a HEPA filter to the sample inlet to remove the particles.  This will determine the sensitivity (noise level) of the instrument.  This test is especially important for instruments that are being installed in pristine environments.


5). Make record of the range of photo-detector current (1873 nA) and make sure that the signal and reference values are in the range of 200,000 to 600,000.


6). PSAP has positive interference from the relative humidity. Ensure either the sample stack has a heated inlet which is relative humidity controlled or install a small heater inside the PSAP so that the inlet is gently heated by a few degrees. (see Comment 1)

7). Calibrate the flow meter in the PSAP by using a proven method.  For example, a Gilibrator (Sensidyne, Gilian) can be used to calibrate the volume flow through the PSAP inlet, but an accurate measure of temperature and pressure at the time of calibration will be required to convert the flow to that at the standard temperature and pressure (STP) of 25oC and 1 atmosphere.  The calibration of the flow meter is necessary to ensure that the volume used in equation (1) and used in the calculation by the firmware is correct.


8). An accurate measure of the spot size of the exposed area on the sample filter is needed in the calculation shown in equation 1.  Use a hand held reticle with scale accurate to 0.05 mm to measure the diameter of the spot size.  Take the average of spot size diameter taken by at least 3 people on at least 10 samples on a monthly basis.


9). It is advised that data should be collected at a highest reasonable time resolution so that fast changes in the aerosol response to absorption are recorded.  However, there is an option to choose the desired time-step for data averaging in the PSAP software.  This could be exercised at the sampling locations depending on how polluted the air-mass is?


10). For a 3W PSAP, go through the PSAP set-up windows and check that initial parameters (default values) with the measured parameters as follows:


If using NOAA software as a data collection platform then adjust the parameters as follows:

The default flow parameters are 1000 for 0, 1240 for 0.3 and 2600 for 2

The parameters A=0.814, B=1.237 and Area should be set to 17.81 mm2.

The spot-size and 5-point flow corrections are applied in the software as data is processed.


If using your own software for data collection then adjust the parameters as follows:

The default flow parameters are 1000 for 0, 1240 for 0.3 and 2600 for 2 and these should be changed to corrected measured flow.

The parameters A=0.814, B=1.237 and Area should be set to the measured value in mm2.

Sample inlet

The sampling set-up, which includes a common aerosol sampling inlet for all instruments, size-cuts and sample conditioning, is described in GAW report # 153 (2003).  According to the report the aerosol sample inlet must be set-up to ensure transmission of particles between 0.1 to 10 µm size ranges.


Determine the operational detection limit for the PSAP as installed in the sampling system.  This can be done by placing a HEPA filter on the entrance of sample intake if possible or in-line to the PSAP and collect data in this orientation for at least 24 hours to determine the changes in the zero absorption.  The detection limit of the instrument is determined as 3 times the standard deviation above the mean absorption measurements.

Data collection

The difference in the data collection between the single wavelength (1W) and the 3W PSAPs is that the 1W PSAP has serial and analog output options but the 3W PSAP has serial output only.  Serial data are transmitted once every second and are easily collected using a computer.  The absorption data are updated at the end of the averaging period selected in the firmware.  Any software capable of capturing a serial data-stream can be used as data collection platform.  The data transmission rate of the PSAP (baud rate=9600 for new PSAPs; some old models require baud rate of 4800), specified in the PSAP setup, needs to be used in the software to collect the data properly.  A data-logger is required for capturing the analog data-stream. Similar to the 1W PSAP, the initial sample and reference intensities are also available in the 3W PSAP data stream but these raw readings are embedded in a hexadecimal format in the end column.  If needed, a program is available from the supplier to extract these raw data for changes to absorption at a later date.

Calibration and Raw Data Adjustments

The filter-based optical methods for the determination of aerosol light absorption coefficients have limitations and artifacts as changes in light attenuation depend on the nature of the filter matrix and light scattering constituents of the aerosols deposited onto the filter.  A calibration method devised by Bond et al. (1999) has been used widely to correct for the filter loading and light scattering artifacts.

Prior to such correction is applied to the data, the raw light absorption coefficient data must be adjusted for the spot size and flow corrections post data collection procedure.  There is a discrepancy in the reported spot size by Bond et al. (1999) to derive their correction scheme and spot size that PSAP firmware uses to calculate the absorption coefficient.  The spot area of 17.83 mm2 is used in the 1W PSAP firmware but the manufacturer’s reference of 20.43 mm2 of area has been used to derive the Bond et al. (1999) correction scheme.  Thus, an adjustment of 0.873 must be applied to compensate for this discrepancy in the areas (Report by Ogren & Andrews).  Equation 3, below, is used to do this adjustment:




where Qpsap is the flow used by PSAP, Qmeas is the flow obtained after calibration, Dmeas is the spot size measured, and Dcalib = 5.1 mm is a reference used in the PSAP calibration by the manufacturer (Bond et al., 1999). However, for the 3W PSAP, there is an option in the set-up menu where the spot area can be preset before measurements are started as mentioned in section 3.1 (10).

  The technique for light scattering corrections suggested by Bond et al (1999), which requires light scattering measurements, should be implemented to correct the PSAP data at all GAW sites as shown in Equation 4:




where σadj is obtained from Equation 3, K1=0.02±0.02 and K2=1.22±0.20 have been derived by Bond et al. (1999), σsp is the light scattering coefficient at 550 nm as measured by a nephelometer.  Bond et al. (1999) suggested that for large scattering aerosols (e.g., sea salt, dust, etc.), the K1 value could be as low as 0.01, although no specific guidance is given on the relationship between particle size and K1.

Data Processing and Archive

The following procedures must be applied to the raw light absorption data as recorded by the acquisition program:


1). When processing raw data stream, it is advisable to take certain steps in maintaining the integrity of the measurements.  The quality control procedures must be followed with proper flagging procedures.  Data must be flagged invalid if one or more of the following conditions are met:

          (i)  there is no flow through the filter due to broken pump or a major leak,

          (ii)  transmittance values decrease below 0.6; the measurements may now have artifacts from aerosol shadowing.

          (iii)  there are spikes in the data due to local contamination from nearby sources.  The contamination sector should be determined.  Aerosol number concentrations could be used, if available, as a measure to determine the local contamination when the winds are from the contamination sector.

          (iv)  when the values for absorption are negative.  This can happen when absorption values are small, and in Equation (1) the transmittance I in the current averaging period is nominally smaller than the I0 in the previous averaging period due to uncertainties in the transmission measurements.


2). Check the time-step to identify any data-gaps in the data-set.


3). Apply the spot size and flow corrections to the light absorption data as described in Section 5.


4). Light absorption data must be merged with light scattering data after the initial quality control process is achieved. Apply Bond et al. (1999) correction for the effect of light scattering, as described in Section 5, to the light absorption data to compensate for scattering by particles. Use the light scattering data at 550 nm.


5). Report hourly and daily averages of light absorption coefficients time series in the final data set as per the guidelines of the GAW, or as required by the local needs.


6). If submitting these data to the World Aerosol Data Center then the archival of these data must follow a proper NARSTO Data Exchange Standard (DES) as adapted by the World Data Center situated in Ispra, Italy.  This DES can be obtained from the World Aerosol Data Center.

Maintenance and Repairs

Daily checks

1) Check that the pump is running and there is flow through the PSAP as set between 1 to 2 standard liter per minute.

2) Check that the filter is placed in the correct orientation with whiter & brighter side up in the filter holder.

3) Check that the filter is changed before the transmittance value reaches 0.6.

4) Check that the o-rings are still in place before and after the filter is changed.

5) While changing the filter, examine the exposed area and ensure that it is sharp and does not appear dispersed or fuzzy at the edges.  If it is dispersed, there is a leak, i.e., the o-rings are not making a good seal.  Examine the o-rings and replace if necessary.

Monthly checks and cleaning

1) Cleaning of the sample cavity is important on a regular basis to maintain low detection limits for the absorption measurements.  The sample cavity must be cleaned once a month with a swab dipped in methanol.  Ensure that fibers from the swab are not left behind in the cavity.  Record the raw signal value on the monthly log sheet before the cleaning by flicking the up switch until you get to the signal and reference window.  Record the signal value after cleaning.  Do the same with reference cavity.

2) Check the performance of the lamp by recording the sample and reference  readings (between 200,000 to 600,000) as well as lamp voltage and current on a regular basis.

3) Inspect the inside of the enclosure box and make sure all tubings are still firmly connected and all electrical connectors are securely in place.

4) Check to see if the o-rings are still in place, determine the integrity of the o-rings and make sure that there are no breaks in the o-rings.  Replace the o-rings if they are missing or damaged.

5) Measure the spot size accurately on 10 filters per month.

Lamp Change

For the 1W PSAP, the lamp is an ultra-brite green LED.  There is a spare taped to the inside of all instrument case.  Remove the rubber cap, and cut or de-solder the leads from the lamp.  The lamp is press fitted with a small amount of glue to hold it in place.  This seal could be broken with a pair of pliers.  Install the new lamp and re-solder the power leads to the lamp leads.  The long lead on the LED is positive and should be connected to the center lead of the shielded wire.  The shield connects to the short LED lead (as described in page 8 of the PSAP manual).


For the 3W PSAP, changing the light source should left to experienced operators or to be performed by the factory.

Table 1 - Daily Check Sheet

Filter ID#
Sample Start


Date     Time


Sample Stop
Date   Time
(end time)
Flow check
O-ring Check
Spot size appearance check

Table 1-Daily Check Sheet

Table 2 - Monthly Maintenance Logsheet

Time (UTC)
Sample Cavity Cleaning
Reference Cavity Cleaning
Lamp Voltage
Lamp Current
Spot size diameter

Table 2-Monthly Maintenance Logsheet


Bond, T.C., T. L. Anderson & D. Campbell (1999), Calibration & intercomparison of filter-based measurements of visible light absorption by aerosols, Aerosol Science & Technology 30, 582-600.

GAW report # 153, 2003.

PSAP manual, Radiance Inc.

Ogren & Andrew, Comment on “Calibration and intercomparison of Filter-Based Measurements of Visible Light Absorption by Aerosols”.

Personal tools
Create a book