Environ Monit Assess (2015) 187: 372 DOI 10.1007/s10661-015-4597-0

Fiber optic light sensor Wayne Chudyk & Kyle F. Flynn

Received: 9 September 2014 / Accepted: 13 May 2015 / Published online: 26 May 2015 # Springer International Publishing Switzerland 2015

Abstract We describe a low-cost fiber optic sensor for measuring photosynthetically active radiation (PAR) in turbulent flow. Existing technology was combined in a novel way for probe development addressing the need for a small but durable instrument for use in flowing water. Optical components including fiber optics and a wide-spectrum light detector were used to separate light collection from electronic detection so that measurements could be completed in either the field or laboratory, in air or underwater. Connection of the detector to Arduino open-source electronics and a portable personal computer (PC) enabled signal processing and allowed data to be stored in a spreadsheet for ease of analysis. Calibration to a commercial cosine-corrected instrument showed suitable agreement with the added benefit that the small sensor face allowed measurements in tight spaces such as close to the streambed or within leafy or filamentous plant growth. Subsequently, we applied the probe in a separate study where over 35 experiments were successfully completed to characterize downward light attenuation in filamentous algae in turbulent flow. Keywords Fiber optic sensor . Photosynthetically active radiation . PAR . Light . Meter . Optics W. Chudyk (*) : K. F. Flynn Civil and Environmental Engineering Department, Tufts University, Medford, MA 02155, USA e-mail: [email protected] K. F. Flynn Montana Department of Environmental Quality, 1520 E. 6th Avenue, Helena, MT 59620, USA

Introduction Measurement of solar radiation is central to determining e n e rg y a v a i l a b i l i t y f o r p h o t o s y n t h e s i s a n d photodegradation. The wavelength range from 400 to 700 nm, or photosynthetically active radiation (PAR), is of primary interest due to its importance to photosynthesizing plants. Thus, the ability to measure PAR both in air and water, including surfaces within centimeters of the channel bottom, is of central importance to determining energy input to ecological systems. Such data are of great interest to biologists and ecologists due to their ecological relevance with regard to primary productivity and ecosystem metabolism (Long et al. 2012; Muchow and Kerven 1977; Ross and Sulev 2000; Pérez et al. 2013; Koch 2001; Kirk 1994) and govern the net autotrophy or heterotrophy of aquatic systems. In streams or rivers, the amount of irradiance (photons) reaching the phytobenthos is influenced by many factors including atmospheric transmission, terrestrial or submerged aquatic vegetation, and water turbidity (Binzer and Sand-Jensen 2002; Hill et al. 2010; Brush and Nixon 2002; Krause-Jensen and Sand-Jensen 1998). These factors can all reduce irradiance in proportion to the absorption and scattering properties of each medium. Consequently, the amount of radiation incident upon the streambed is less than received by terrestrial plants, but still sufficient to satisfy photobiological processes. The effect of light availability on the photosynthetic response can also indirectly affect organisms dependent on the phytobenthos (Law 2011; Ohta et al. 2011; Żbikowski and Kobak 2007; Sturt et al. 2011).

372 Page 2 of 7

PAR in the aquatic environment can be separated into upward and downward components or, alternatively, spherically integrated scalar irradiance (Kirk 1994). Downwelling irradiance is most commonly used for quantification of light intensity in environmental microbiology (Kühl 2005), though optical studies with field radiance microprobes show that diffuse (scattered) light can be a major source of energy for biofilms or sediments (Kühl 2005). For downwelling irradiance, cosine collectors are often recommended to measure PAR in highly turbid waters or at depths of tens of meters (Jerlov 1968; Preisendorfer and Mobley 1984; Kirk 1991; Kirk 2003). In shallow clear waters flat-plate collectors are appropriate (Højerslev 1975). In addition, Secchi disk, nephelometric turbidity, and underwater light meters have been found to be useful in estimating or directly measuring light penetration into the water of both lotic (flowing) and lentic (lake) systems (Pérez et al. 2013; Carter and Rybicki 1990; Civera et al. 2011; Long et al. 2012). PAR measurements are commonly done with commercial instruments (Kirk 1994), or custom-built optical sensors, that differentiate light propagating through either the water column, foliage, or other materials that scatter or absorb photons. PAR can also be measured indirectly using photodegrading dyes (Bechtold et al. 2012; Long et al. 2012). Hence, the ability to measure PAR both in air and water, especially at fine resolution at or near the benthic substrate, is of great ecological importance. A major disadvantage of many commercial optical instruments is that they are larger, or have upward projecting sensors, that are difficult to use when faced with making measurements close to horizontal surfaces or small spaces such as within aquatic or terrestrial plants or near the substrate (rocks) of a streambed. A majority have a minimum vertical dimension of 2 to 5 cm, though it is noted that the commercialgrade Ocean Optics sensor has a profile height of approximately 4 cm when oriented vertically (part numbers CC-3-UV-S, 74-UV, and 74-90-UV; data from Ocean Optics); it is much more expensive. Many others have a cable connection on the bottom, or are cylindrical in shape, with diameters >2.5 cm (Civera et al. 2011; LI-COR 2011; Benham and George 1981; Fielder and Comeau 2000; Booth 1976; Burr and Burr 1934; Melbourne and Daniel 2003; Pontailler 1990). In this regard, most commercial sensors are difficult to apply under certain circumstances. They also can be prohibitive due to cost.

Environ Monit Assess (2015) 187: 372

The propensity of larger sensors to influence their measurement region is another concern when making in situ measurements. For example, relatively large sensors can affect the fluid motion of media (air or water), which can sometimes lead to erroneous measurements (Fox et al. 2009). As an example, PAR is affected by changes in the flow regime of air or water when flexible plant parts are present due to movement of the plant material during turbulent fluctuations (Nepf and Ghisalberti 2008; Wilson et al. 2003). Furthermore, macrophytes or sea grasses are often located near the bottom of the streambed and closely shade horizontal surfaces. PAR measurements in these locations are difficult to obtain as plant components can intermittently block light when moving back and forth in the flow. Consequently, accurate measurements of light transmission in periphyton or macrophytes should minimize the changes in the environment in which they are being conducted, be affordable, and, if possible, be completed in situ (Frankovich and Zieman 2005, citing others). Bare or shaped optical fiber sensors sometimes are used to minimize the influence of disturbances of flow or structure induced by larger-sized sensors (Jørgensen and Des Marais 1986; Dodds 1992; Kühl et al. 1997; Kühl 2005). While these have been very useful in microscale and laboratory studies, fragility has limited their use in field or flume settings as optical fibers are at risk for stress fracture or breakage as a result of water motion. Small unsupported fibers are also difficult to position accurately and reproducibly in moving water. As a consequence, a durable, yet small, affordable instrument for measuring in situ PAR is needed and described here.

Materials and methods Sensor description To meet our objective of providing a durable but small sensor, a combination of fiber optic components was used. Downwelling light was collected using a vertically oriented 12.5-mm BK 7 right-angle prism (Edmund Optics) attached to a SMA fiber adapter plate (Thorlabs) with optically clear epoxy (Epoxyset), which allows measurement 13 mm from the substrate (compared to >40 mm for Ocean Optics). A 1-m-long 1000-μm core fiber optic patch cord (LEONI Fiber Optics), with SMA 905 connectors at each end, transmitted light from the prism to the detector. Optionally, to

Environ Monit Assess (2015) 187: 372

Page 3 of 7 372

obtain measurements of light incident other than vertically (e.g., diffuse with cosine correction), a Plexiglas diffuser plate with surfaces made matt by abrasion after Dawson (1980) was attached to the prism top with epoxy. Either sensor can be useful for measuring different components of downwelling irradiance. Inside the detector, a SMA fiber adapter plate was epoxied to a 400-nm long-pass optical filter (Edmund Optics) which was then stacked onto a 700-nm shortpass optical filter (Edmund Optics). Finally, this was affixed over a general-purpose optical (photodiode) detector chip (TSL230RD) on a prototyping breakout board (Adafruit Proto Shield for Arduino) compatible with an Arduino Uno R3 (Adafruit Industries) controller. The Arduino is a widely used open-source platform that was programmed to export data to a portable PC through a USB connection with modification of software presented by others (Parallax Inc. 2014; Thomson 2010). The setup described above resulted in total light intensity 400–700 nm expressed as a frequency count recorded using Microsoft Excel at 1-s increments, which were later averaged over 30–60-s periods, depending on application. The overall cost of the sensor and detector was approximately $350. The arrangement of light sensor components is shown in Fig. 1.

Transmission measurements at different wavelengths Transmission measurements of optical filters were obtained using a Varian Cary 50 Bio UV-visible spectrophotometer. The optical filters were placed in the sample chamber of the spectrophotometer so that they were in the instrument’s light path, and percent transmission was recorded at wavelength intervals of 1 nm. The overall response of the optical sensor was then determined by combining data from spectrophotometermeasured transmission of the combined optical filters with the photodiode spectral responsivity reported by the manufacturer (Taos 2007). Wire to PC

Incident light

Incident light and cosine response Evaluation of cosine response to non-incident light according to Dawson (1980) was done using the prism assembly coupled to a CVI Spectral Products SM-240 spectrometer via a fiber optic patch cord. The light source for non-incident light measurement was an LED with 640-nm peak output (Nichia) coupled to a fiber optic patch cord. LED output was oriented so that a 1-cm spot was directed onto the surface of the prism, and then, the prism was rotated through 10° increments, where light intensity at 640 nm was measured by the spectrometer. The acceptance angle of the sensor was calculated using the numerical aperture of the fiber optic cable used to collect light from the prism (Kühl 2005; Crisp and Elliot 2005). NA ¼ n sinΘa     NA 0:22 ¼ sin−1 ¼ sin−1 ð0:165Þ ¼ 9:5∘ Θa ¼ sin−1 n 1:33

where Θa is the acceptance angle, NA is the numerical aperture of the optical fiber, and n is the refractive index of the optical fiber’s silica core. Sensor calibration Calibration of the PAR sensor to correct between sensor output frequency (10−5) and PAR (μE m−2 s−1) was completed in natural sunlight using a LI-COR 192 (Lincoln, NB) cosine-corrected meter at zenith angle 25° at solar noon without the diffuser, a time when field measurements were to be conducted. Calibration was done in a black non-reflective container filled with tap water, using a suspension of clay particles to limit light penetration into the water. A secondary calibration was completed at an angle of 0° under metal halide lamp (Hortilux Blue 1000 W) to differentiate the importance of using a unique calibration at various incident angles (θ).

Results

Light filters

Detector response versus wavelength prism

Fiber opc cable

Detector

Fig. 1 Arrangement of light sensor components allowing placement of the small prism and fiber optic cable within filamentous plants to measure light attenuation at convenient locations and close to surfaces such as the bottom of the streambed

The individual transmission and combined normalized detector response versus wavelength for the two optical filters and optical sensor responsivity are shown in Fig. 2. Overall, the combined filters provide a sharp and

372 Page 4 of 7

Environ Monit Assess (2015) 187: 372

Fig. 2 Normalized detector response as a function of wavelength, combining optical filter transmissivity and photodiode spectral responsivity

desirable cutoff at 400 and 700 nm, but with a uniform response (80–90 %) between the wavelengths. The photodiode, however, is more wavelength dependent, with low responsivity at 400 nm (40 %) and nearly double that at 700 nm (80 %). Thus, the combined response over the PAR region is shaped primarily by the photodiode with some attenuation by the optical filters. The combined response of the detector and optical filters is shown again in Fig. 3, but this time, against the ideal response for a quantum-based PAR detector (according to quantum efficiency). The sensor approaches ideal response, where energy is inversely proportional to wavelength (Aaslyng et al. 1999; Dawson 1980; Fielder and Comeau 2000; Jones et al. 2003; McPherson 1969). Incident light and cosine response

drop-off in irradiance response occurs between 10 and 40°, where only 20 % of the incident light is available at 40°. Use of a diffuser attached to the top surface of the prism greatly improves the measurement of light incident away from vertical (Fig. 4) and very nearly follows the cosine response (Dawson 1980; Aaslyng et al. 1999). However, use of the diffuser has its drawbacks, causing blockage of approximately 45 % incident light and subsequent signal loss (not shown). In this regard, if only downwelling light levels are of interest (±10 of zero degrees incidence), the diffuser can be left off provided that an appropriate calibration factor can be established for a given incident angle (θ) at which the measurement is being made. If low PAR levels are expected, the diffuser may impede the measurement and reduce the overall sensitivity beyond unacceptable levels.

As described in the BMaterials and methods^ section, light losses are expected at angles greater than 9.5° from vertical as demonstrated experimentally in Fig. 4. A sharp

Fig. 3 Normalized detector response as a function of wavelength compared with ideal response for a quantum-based PAR detector (Dawson 1980)

Fig. 4 Sensor response versus incident light angle, with diffuser (triangles) and without diffuser (squares). The ideal cosine response (Aaslyng et al. 1999) is the dotted line

Environ Monit Assess (2015) 187: 372

Page 5 of 7 372

Calibration to standardized meter

recirculating racetrack flume and also in situ in the Clark Fork River, USA (Flynn 2014; Flynn et al. 2014). While the work described here is primarily applicable to aquatic systems, it is important to point out that the sensor is pertinent to other environments such as landbased studies. An example would be measurement of light propagation through dense or wind-stirred foliage or more generalized PAR measurement. Using longer or shorter optical fibers will allow application in a variety of locations. However, use of this sensor requires appropriate calibration. With the diffuser in place, indirect (cosine corrected) light losses were about 45 % due to transmission through the diffuser. Additionally, under natural sunlight conditions where incident angles are often greater than zero, an angle-specific calibration was needed. As seen in this study, for an incident angle of 25°, about half incident light is not measured (Fig. 4). Care must therefore be taken to calibrate the sensor under conditions consistent with the zenith for which incident downwelling light is being measured.

The calibration to a standardized meter (Fig. 5) shows good agreement (r2 =0.98) between sensor response and a LI-COR 192 in natural sunlight at an incidence angle of 25°. Thus, the sensor can be readily calibrated to quantify PAR. However, for measurements at different θ values, a new calibration coefficient must be acquired for each θ to correct between sensor output frequency and PAR. In other words, a unique numerical correction is required to translate between the sensor without the diffuser and the idealized cosine response due to the relationship between percent incident light and incident angle shown previously in Fig. 4.

Discussion The sensor developed in this study is useful where underwater measurements are needed such as near the streambed, in filamentous algae, or other difficult settings such as in turbulent flow where commercial sensors may not be able to reach or are not affordable. The use of standard parts, durable connectors, and opensource equipment provides a methodology for construction of a small-size durable PAR sensor at reasonable cost that provides comparable results to standardized meters. In addition, use of the Arduino electronic components, a single photodiode, and cutoff filters could be adapted to many other fiber optic setups including the measurement of spherical irradiance (Dodds 1992) or other upwelling radiation. In this case, a number of measurements have been obtained for simultaneous downwelling (±10° from incident angle) profiles through algal mats with both the sensor and LI-COR in such settings with reasonable agreement in a

Fig. 5 Sensor calibration versus LI-COR 192 in natural sunlight attenuated by a suspension of clay particles at an incidence angle of 25°. Based on an averaging period of 30 s for each observation

Conclusions The sensor described in this paper allows irradiance to be measured in difficult conditions such as in turbulent flow. Additionally, its small size and relative durability allow it to be placed reproducibly with a micromanipulator in locations that may be difficult to access with a commercial meter, such as near the streambed. Accordingly, it has merit for cases where commercial sensors are prohibitive due either to physical size, cost, or both. Directions for further research include the following: 1. Evaluation of limits to sensor performance such as length of the optical fiber. Long fibers would allow measurements within streams of various widths and depths. 2. Light measurements in lentic systems. Use of the sensor in moving water has already been demonstrated; however, there is potential for similar measurements in ponds and lakes. 3. Use of this sensor in further studies of algal biomass effects on PAR extinction or inclusion of shading in mathematical models of riverine systems. 4. Testing in non-aqueous environments. Data could be obtained on land within dense moving foliage with this sensor to determine shading effects in a variety of terrestrial environments.

372 Page 6 of 7 Acknowledgments The authors wish to thank the MontanaWyoming USGS Science Center for loan of their Li-COR 192 m used in support of this work. The authors also wish to thank Kevin Chudyk for layout, assembly, and testing of the photodiode detector breakout board circuitry. Portions of this work were presented at SPIE Photonics Europe Innovation Village April 13–18, 2014.

References Aaslyng, J. M., Rosenqvist, E., & Høgh-Schmidt, K. (1999). A sensor for microclimatic measurement of photosynthetically active radiation in a plant canopy. Agricultural and Forest Meteorology, 96, 189–197. doi:10.1016/S0168-1923(99)00057-X. Bechtold, H. A., Rosi-Marshall, E. J., Warren, D. R., & Cole, J. J. (2012). A practical method for measuring integrated solar radiation reaching streambeds using photodegrading dyes. Freshwater Sciences, 31, 1070–1077. doi:10.1899/12-003.1. Benham, D. G., & George, D. G. (1981). A portable system for measuring water temperature, conductivity, dissolved oxygen, light attenuation and depth of sampling. Freshwater Biology, 11, 459–471. doi:10.1111/j.1365-2427.1981. tb01277.x. Binzer, T., & Sand-Jensen, K. (2002). Production in aquatic macrophyte communities: a theoretical and empirical study of the influence of spatial light distribution. Limnology and Oceanography, 47, 1742–1750. ISSN: 0024-3590. Booth, C. R. (1976). The design and evaluation of a measurement system for photosynthetically active quantum scalar irradiance. Limnology and Oceanography, 21, 326–336. ISSN: 0024–3590. Brush, M. J., & Nixon, S. W. (2002). Direct measurements of light attenuation by epiphytes on eelgrass Zostera marina. Marine Ecology Progress Series, 238, 73–79. doi:10.3354/ meps238073. Burr, G. O., & Burr, M. M. (1934). A rapid survey instrument for the measurement of light intensity under water. Ecology, 15, 326–328. doi:10.2307/1932481. Carter, V., & Rybicki, N. B. (1990). Light attenuation and submersed macrophyte distribution in the tidal Potomac River and estuary. Estuaries, 13, 441–452. doi:10.2307/1351788. Civera, J. I., Gil, R. I., Laguarda-Miro, N., Garcia-Breijo, E., GilSánchez, & Martínez-Guijarro, R. (2011). Instrument for sunlight extinction measurement in water bodies. Sensors and Actuators A, 168, 267–274. doi:10.1016/j.sna.2011.04.025. Crisp, J., & Elliot, B. (2005). Introduction to fiber optics (3rd ed.). Boston: Elsevier. Dawson, F. H. (1980). An inexpensive photosynthetic irradiance sensor for ecological field studies. Hydrobiologia, 77, 71–76. doi:10.1007/BF00006390. Dodds, W. K. (1992). A modified fiber-optic light microprobe to measure spherically integrated photosynthetic photon flux density: characterization of periphyton photosynthesis-irradiance patterns. Limnology and Oceanography, 37, 871–878. doi:10.4319/lo.1992.37.4. 0871. Fielder, P. & Comeau, P. (2000). Construction and testing of an inexpensive PAR sensor. Research Branch British Columbia

Environ Monit Assess (2015) 187: 372 Ministry of Forests Working paper 53, Victoria, B.C. ISBN: 0-7726-4392-X. Flynn, K. F. (2014). Methods and mathematical approaches for modeling Cladophora glomerata and river periphyton, Ph.D. thesis, Dept. of Civil & Env. Eng., Tufts University, Medford, MA. Flynn K. F., Chudyk, W. A., Chapra, S. C., & Watson, V. (2014). Influence of biomass and velocity on light attenuation of Cladophora glomerata L. (Kuetzing) in rivers, in preparation. Fox, R. W., Pritchard, P. J., & McDonald, A. T. (2009). Introduction to fluid mechanics (7th ed.). Hoboken: Wiley. ISBN 9780471742999. Frankovich, T. A., & Zieman, J. C. (2005). Periphyton light transmission relationships in Florida Bay and the Florida Keys, USA. Aquatic Botany, 83, 14–30. doi:10.1016/j. aquabot.2005.05.003. Hill, W. R., Smith, J. G., & Stewart, A. J. (2010). Light, nutrients, and herbivore growth in oligotrophic streams. Ecology, 91, 518–527. doi:10.1890/09-0703.1. Højerslev, N. (1975). A spectral light absorption meter for measurements in the sea. Limnology and Oceanography, 20, 1024–1034. ISSN: 0024-3590. Jerlov, N. G. (1968). Optical oceanography. New York: Elsevier Publishing Company. ISBN 978-0444403209. Jones, H. G., Archer, N., Rotenberg, E., & Casa, R. (2003). Radiation measurement for plant ecophysiology. Journal of Experimental Botany, 54, 879–889. doi:10.1093/jxb/erg116. Jørgensen, B. B., & Des Marais, D. J. (1986). A simple fiber-optic microprobe for high resolution light measurements: application in marine sediment. Limnology and Oceanography, 31, 1376–1383. doi:10.4319/lo.1986.31.6.1376. Kirk, J. T. O. (1991). Volume scattering function, average cosines, and the underwater light field. Limnology and Oceanography, 36, 455–467. doi:10.4319/lo.1991.36.3.0455. Kirk, J. T. O. (1994). Light and photosynthesis in aquatic ecosystems. New York: Cambridge University Press. ISBN 9780511623370. Kirk, J. T. O. (2003). The vertical attenuation of irradiance as a function of the optical properties of the water. Limnology and Oceanography, 48, 9–17. doi:10.4319/lo.2003.48.1.0009. Koch, E. W. (2001). Beyond light: physical, geological, and geochemical parameters as possible submersed aquatic vegetation habitat requirements. Estuaries, 24, 1–17. doi:10. 2307/1352808. Krause-Jensen, D., & Sand-Jensen, K. (1998). Light attenuation and photosynthesis of aquatic plant communities. Limnology and Oceanography, 43, 396–407. ISSN: 0024–3590. Kühl, M. (2005). Optical microsensors for analysis of microbial communities. Methods in Enzymology, 397, 166–199. doi: 10.1016/S0076-6879(05)97010-9. Kühl, M., Lassen, C., & Revsbech, N. P. (1997). A simple light meter for measurements of PAR (400 to 700 nm) with fiberoptic microprobes: application for P vs E-0(PAR) measurements in a microbial mat. Aquatic Microbial Ecology, 13, 197–207. doi:10.3354/ame013197. Law, R. (2011). A review of the function and uses of, and factors affecting, stream phytobenthos. Freshwater Reviews, 4, 135– 166. doi:10.1608/FRJ-4.1.448. LI-COR (2011). Radiation measurement instruments, www.licor. com. Accessed 2/14/2014.

Environ Monit Assess (2015) 187: 372 Long, M. H., Rheuban, J. E., Berg, P., & Zieman, J. C. (2012). A comparison and correction of light intensity loggers to photosynthetically active radiation sensors. Limnology and Oceanography: Methods, 10, 416–424. doi:10.4319/lom. 2012.10.416. McPherson, H. G. (1969). Photocell-filter combinations for measuring photosynthetically active radiation. Agricultural Meteorology, 6, 347–356. doi:10.1016/0002-1571(69) 90026-0. Melbourne, B. A., & Daniel, P. J. (2003). A low-cost sensor for measuring spatiotemporal variation of light intensity on the streambed. Journal of the North American Benthological Society, 22, 143–151. doi:10.2307/1467983. Muchow, R. C., & Kerven, G. L. (1977). A low cost instrument for measurement of photosynthetically active radiation in field canopies. Agricultural Meteorology, 18, 187–195. doi:10. 1016/0002-1571(77)90036-X. Nepf, H., & Ghisalberti, M. (2008). Flow and transport in channels with submerged vegetation. Acta Geophysica, 56, 753–777. doi:10.2478/s11600-008-0017-y. Ohta, T., Miyake, Y., & Hiura, T. (2011). Light intensity regulates growth and reproduction of a snail grazer (Gyraulus chinensis) through changes in the quality and biomass of stream periphyton. Freshwater Biology, 56, 2260–2271. doi:10.1111/j.1365-2427.2011.02653.x. Parallax, Inc. (2014). PLX-DAQ, http://www.parallax.com/ downloads/plx-daq. Accessed 21 Feb 2014. Pérez, G. L., Lagomarsino, L., & Zagarese, H. E. (2013). Optical properties of highly turbid shallow lakes with contrasting turbidity origins: the ecological and water management

Page 7 of 7 372 implications. Journal of Environmental Management, 130, 207–220. doi:10.1016/j.jenvman.2013.09.001. Pontailler, J. (1990). A cheap quantum sensor using a gallium arsenide photodiode. Functional Ecology, 4, 591–596. doi: 10.2307/2389327. Preisendorfer, R. W., & Mobley, C. D. (1984). Direct and inverse irradiance models in hydrologic optics. Limnology and Oceanography, 29, 903–929. doi:10.4319/lo.1984.29.5.0903. Ross, J., & Sulev, M. (2000). Sources of errors in measurements of PAR. Agricultural and Forest Meteorology, 100, 103–125. doi:10.1016/S0168-1923(99)00144-6. Sturt, M. M., Jansen, M. A. K., & Harrison, S. S. C. (2011). Invertebrate grazing and riparian shade as controllers of nuisance algae in a eutrophic river. Freshwater Biology, 56, 2580–2593. doi:10.1111/j.1365-2427.2011.02684.x. Taos (2007). TSL230RD, TSL230ARD, TSL230BRD Programmable light-to-frequency converters, www.taosinc. com. Accessed 9/7/2012. Thomson, J. (2010). Getting started with the TSL230R and Arduino, http://jethomson.wordpress.com/tsl230r-articles/getting-startedwith-the-tsl230r-and-arduino. Accessed 21 Feb 2014. Wilson, C. A. M. E., Stoesser, T., Bates, P. D., & Pinzen, A. B. (2003). Open channel flow through different forms of submerged flexible vegetation. Journal of Hydraulic Engineering, 129, 847–853. doi:10.1061/∼ASCE!07339429∼2003!129:11∼847!. Żbikowski, J., & Kobak, J. (2007). Factors influencing taxonomic composition and abundance of macrozoobenthos in extralittoral zone of shallow eutrophic lakes. Hydrobiologia, 584, 145–155. doi:10.1007/s10750-007-0613-x.

Fiber optic light sensor.

We describe a low-cost fiber optic sensor for measuring photosynthetically active radiation (PAR) in turbulent flow. Existing technology was combined ...
462KB Sizes 1 Downloads 13 Views