JP4719915B2 - Method for estimating soil-derived suspension concentration in inland water - Google Patents

Description

  The present invention relates to a method for estimating the concentration of suspended matter derived from soil in rivers and lakes.

  The water environment such as rivers and lakes is one of the important elements in the living environment including drinking water and in fishery resources. Water quality monitoring technology, which is a prerequisite for improving water quality, is becoming increasingly important. Therefore, a highly reliable water quality measuring means is required.

  Conventional water quality measurement techniques can be categorized from the viewpoint of manual operation, and those that are performed by technicians using measuring instruments, etc. on the spot, and samples taken on-site are brought back to the laboratory and measured with measuring instruments. There is a thing.

  The former is, for example, on-site measurement of water appearance, smell, water temperature, pH, ORP, electrical conductivity, transparency, dissolved oxygen, CODmn, etc., and if there is a measuring instrument, the measurement result can be obtained in a short time. However, there are drawbacks that require technicians and costly periodic measurements for monitoring.

  The latter can be measured in the laboratory using precision measuring equipment corresponding to the measurement items such as water soil or biological suspension concentration, BOD, CODcr, or nitrogen concentration of ammonia, nitrous acid, nitric acid, etc. However, it is necessary to take the sample collected at the site back to the laboratory in a form that does not change, and it is difficult to regularly analyze and monitor the water quality at remote locations.

  In any case, the method requires several days to several tens of days after sampling to collect water quality data from these measurement results. In view of this, in some cases, monitoring data is collected continuously or periodically by installing a water quality measurement device using a photoelectric sensor at the site.

  In the conventional soil-derived suspension concentration estimation method, the contact measurement method is limited to a method in which water is collected and analyzed, or a measurement device is directly laid in water. As a non-contact measurement method, in the remote sensing field, a method using the intensity of reflected light from the sky, especially using the intensity of multiple reflected lights in the visible and near-infrared range of 400-1050 nm (See, for example, Patent Document 1, Non-Patent Document 1, and Non-Patent Document 2.)

  For medium-scale rivers with a depth of about 0.5 to 5m, the concentration of the soil-derived suspension in the river surface is estimated by measuring the intensity of reflected light from the sky, etc. (See, for example, Non-Patent Document 3). This method is effective when there is a steady water flow with little difference in soil-derived suspension concentration between the surface layer and the bottom layer. Limited.

  As described above, according to the conventional contact measurement method, it is difficult to measure quickly and multipoint, and the method using reflected light cannot sufficiently cope with the variety of conditions such as the direction of observation and the color of the bottom of the water. It was.

JP 2004-150916 A Yasuharu Yamada, Motoya Saito, Takehiko Okuyama “Spectral Reflectance Measurement of Turbid Water: Case Study of Okinawa Kunigami Merging” Annual Meeting of the Agricultural Civil Society of Japan, p468-469, 1992. J. A. Warrick, L. A. K. Mertes, D. A. Siegel and C. Mackenzie, “Estimating suspended sediment concentrations in turbid coastal waters of the Santa Barbara Channel with SeaWiFS”, International Journal of Remote Sensing, 25, p1995-2002, 2004. Kenji Sakanishi, Yuka Sasaki, Ken-ichi Kanda, Yasuhiro Nakajima, “Understanding the Actual Conditions of Suspended Materials Effluent from Agricultural Lands in a Medium-Scale Basin” Abstracts of Proceedings of the 2004 Annual Meeting of the Agricultural Civil Society of Japan p636-637, 2004.

  This invention makes it a subject to provide the non-contact estimation method which is hard to be influenced by the difference of an observation direction, a bottom color, etc. in the measurement of the soil-derived suspension density | concentration in land water.

The present inventors have removed the polarization component from the reflected light of one wavelength band included in the short-wavelength infrared region, one wavelength band included in the near-infrared region, and one wavelength band included in the visible region. The inventors have found that the measurement error due to the observation direction is reduced and the estimation accuracy of the soil-derived suspension concentration is improved, and the present invention has been achieved. That is, the present invention is as follows.
<1> From the reflected light of sunlight on the surface of land water, one wavelength band included in the short-wavelength infrared region, one wavelength band included in the near-infrared region, and one wavelength band included in the visible region It is a method for estimating the concentration of suspended matter derived from soil in land water by measuring and calculating the polarization reflectance, and the non-polarized reflectance is a polarized light whose reflected light in one wavelength band rotates around the optical axis. It is preferably calculated from the polarization reflection intensity in the wavelength band reflected light detected by passing through the filter.
<2> Further, in the present invention, the arithmetic processing is performed as follows:
SS = A + D + α × NPRs + β × NPRn + γ × NPRv
It is an estimation method of the soil-derived suspension concentration.
In the above calculation process, SS is the concentration of suspended soil derived from land water, D is a correction value due to the difference in bottom color, NPRs is the non-polarized reflectance of one wavelength band s included in the short wavelength infrared region, and NPRn is Non-polarized reflectance of one wavelength band n included in the near infrared region, NPRv is an unpolarized reflectance of one wavelength band v included in the visible region, and α, β, γ, and A are preset constants. Each is shown.
<3> Furthermore, the present invention relates to an aperture unit for focusing light to be measured within a certain visual field range, a filter for selecting a wavelength, a polarizing filter, a polarizing filter rotating mechanism, and a photoelectric converter for photoelectrically converting reflected light. It is a polarization reflection intensity measuring device including a conversion sensor and a voltage output device that amplifies the photoelectrically converted current and then performs synchronous detection and smoothing to extract an output voltage.

  According to the present invention, the concentration of the soil-derived suspension in land water can be accurately estimated regardless of the observation direction, and the difference in the bottom color can be adjusted by increasing or decreasing one coefficient, which is extremely labor-saving.

  The present invention relates to the reflected light of sunlight on the surface of land water, one wavelength band included in the short wavelength infrared region, one wavelength band included in the near infrared region, and one wavelength band included in the visible region. This is an estimation method of the concentration of suspended matter derived from soil in land water (mg / L, hereinafter sometimes referred to as SS) by measuring non-polarized reflectance and performing arithmetic processing. Embodiments of the present invention will be described below.

(Spectra to specific wavelength band)
The reflected light of sunlight on the surface of inland water is narrowed down to a certain visual field range by the aperture part. The aperture section is preferably an aperture angle switching system that can adjust the visual field range according to the distance from the surface of land water. The opening angle switching range of the opening angle switching method is preferably a range where the opening angle half width is 2 ° to 10 °. In the opening angle switching method, the distance between the aperture portion and the surface of land water is preferably in the range of 0.5 to 20 m.

  The reflected light incident on the aperture part is divided into specific wavelength bands by a band pass filter. As a spectral method for each wavelength, it is preferable to use a band-pass filter, and a method using a diffraction grating is not preferable because polarized light is generated at the time of spectroscopy.

  As the wavelength band, a wavelength band included in the short wavelength infrared region, the near infrared region, and the visible region is selected. As a wavelength band to be selected, a wavelength range of 1500 to 2300 nm is preferable in the short wavelength infrared region, a wavelength range of 750 to 1300 nm is preferable in the near infrared region, and a wavelength range of 450 to 700 nm is preferable in the visible range, and a short wavelength is particularly preferable. In the infrared region, a wavelength range of 1600 to 1700 nm is preferable, in the near infrared region, a wavelength range of 780 to 880 nm is more preferable, and in the visible range, a wavelength range of 510 to 610 nm is more preferable.

  The reflection intensity of the reflected light according to wavelength is measured by an output value (hereinafter referred to as a photoelectric conversion value) from a photodetector that performs photoelectric conversion. The reflection intensity of the reflected light according to wavelength is continuously measured as a photoelectric conversion value by a polarization spectroradiometer. By multiplying the photoelectric conversion value by the conversion factor K, the reflection intensity can be obtained.

(Polarization measurement)
The wavelength-specific reflected light that has been dispersed for each wavelength band as described above passes through a polarizing filter that rotates about the optical axis as a central axis, whereby the proportion of polarized light in the wavelength-specific reflected light (hereinafter referred to as the degree of polarization for each wavelength). The amount of passing light increases or decreases in accordance with. The increase / decrease in the amount of light is continuously measured as a photoelectric conversion value for each wavelength by a polarization spectroradiometer, and by multiplying the conversion factor K, the reflection intensity of the polarization component can be obtained. Note that light other than the polarization component in the reflected light is non-polarized reflected light.

  It is preferable that the polarizing filter is continuously rotated around a rotation center axis parallel to the optical axis at a constant speed in a constant direction. The rotation speed is preferably 1 to 3 seconds per rotation, and more preferably 1.5 to 2.5 seconds. The polarization angle measurement range of the polarizing filter needs to be at least 0 ° to 180 °, preferably 0 ° to 360 °. The polarization angle accuracy is preferably in the range of -2 to + 2 °, and more preferably in the range of -1 to + 1 °.

  Furthermore, it is preferable that the polarization filter is corrected for deviation in light transmittance due to manufacturing unevenness of the filter part. In this correction, it is preferable that a calibration value is created in advance using light that does not include a polarization component obtained by making sunlight enter the multiple integrating sphere and performing multiple reflection.

  Since the present invention uses sunlight, an estimated value S for each wavelength of sunlight is used, and the “estimated value S” is calculated from the latitude and longitude of the measurement point and the measurement date and time in the “solar constant”. Calculated as a value multiplied by the cosine of the solar zenith angle. The polarization spectroradiometer preferably includes a zenith angle detection mechanism for the optical axis at the time of measurement and an azimuth detection mechanism based on magnetic north.

(Calculation of polarization degree by wavelength and reflectance by non-polarization wavelength)
For the photoelectric conversion values by wavelength, the maximum photoelectric conversion value (Max) and the minimum photoelectric conversion value (Min) are detected from the photoelectric conversion values obtained for each rotation angle.

  From the photoelectric conversion maximum value (Max) and photoelectric conversion minimum value (Min), the polarization degree P for each wavelength is calculated by the following formula 1. The degree of polarization P for each wavelength is the ratio of the polarization reflection intensity for each wavelength in the reflected light from the object.

    P = (Max−Min) / (Max + Min) (Formula 1)

  Using the photoelectric conversion maximum value (Max), photoelectric conversion minimum value (Min), conversion coefficient K, and the estimated amount S of sunlight for each wavelength, the reflectance R (%) for each wavelength is calculated by the following equation 2.

    R = K × (Max + Min) / (2 × S) × 100 (Formula 2)

  From the polarization degree P by wavelength and the reflectance R by wavelength, the reflectance NPR (%) by non-polarized wavelength is calculated by the following formula 3.

    NPR = R × (1-P) (Formula 3)

  Non-polarized reflectance NPRs of one wavelength band s included in the short-wavelength infrared region obtained by the above, non-polarized reflectance NPRn of one wavelength band n included in the near-infrared region, one wavelength included in the visible region By using the non-polarized reflectance NPRv of band v and the preset constants α, β, γ, and A, and the correction value D due to the difference in the bottom color, the soil-derived suspension concentration SS in land water is It can be obtained by the following formula.

    SS = A + D + α × NPRs + β × NPRn + γ × NPRv (Formula 4)

Hereinafter, although the concrete content of the present invention is explained by an example, the present invention is not limited to the following examples.
(measuring device)
For the measurement of the polarization reflection intensity used in this example, a polarization spectroradiometer composed of a sensor head unit, a control unit, a computer, and a battery power source unit was used. The sensor head unit is an aperture unit that incorporates the light to be measured from the outside within a certain visual field range (incorporated into a lens barrel by Donalec Co., Ltd., and the opening angle half-width 5 ° / 7.5 ° switching type), Polarizing filter (Luceo Co., Ltd., visible range POLAX-42S, and near infrared range POLAX-381RII, cut by Donarec Co., Ltd. and incorporated into a rotating ring processed by the company, polarization angle measurement range: 0-359 °, polarization angle measurement (Accuracy: ± 1 °), and their rotation mechanism (Donnarec Co., Ltd., the above rotating ring is driven by a DC motor, polarizing filter rotation speed: 1 rotation / 2 seconds), band pass filter that divides measurement light by wavelength (center) Wavelength, 490nm, 560nm, 660nm in the visible range, center wavelength in the near infrared range, 830nm, 1150nm, 1250nm, 3 bands, 1650nm, 2200nm, 2 bands from the short wavelength infrared range, 8 bands in total Also A photo detector that photoelectrically converts the amount of light for each wavelength band (manufactured by Shin Kogyo Kogyo Co., Ltd.) (Silicon photo sensor S2386-18K x4 from Hamamatsu Photonics, Germanium photo sensor B2538-01 x3) PbS photosensor P2532 × 1)

  The polarization direction was set to 0 ° vertically upward with respect to the measurement optical axis. This sensor head further includes a zenith angle detection mechanism for the optical axis during measurement (observation zenith angle measurement method: suspension weight and pulse encoder, OMRON E6C2-A, observation zenith angle measurement range: 0 to 180 °, 1 ° Step (0 ° downward, 90 ° horizontal), observation zenith angle measurement accuracy: ± 2 °), and azimuth detection mechanism based on magnetic north (azimuth angle measurement method: magnetic azimuth sensor, AP-311 manufactured by Aplus Co., Ltd.) Azimuth measurement range: 0 to 359 °, 1 ° step, azimuth measurement accuracy: ± 8 °).

  The control unit (manufactured by Donalec Co., Ltd., in-house made of commercially available semiconductors, resistors, capacitors, etc.) converts the amount of light for each wavelength band incident on the sensor head into a voltage value (hereinafter referred to as photoelectric conversion by wavelength). Value)), transmitted to the attached computer, processed and recorded. The photoelectric conversion value for each wavelength is output in the range of 0 to 359 ° for each 1 ° of angle accompanying the rotation of the polarizing filter.

(Correction method for manufacturing unevenness of polarizing filter)
Regarding the deviation of the light transmittance due to the manufacturing unevenness of the filter part due to the rotation of the polarizing filter, a simple integrating sphere (made by Donalec Co., Ltd.) where the inside of a commercially available polystyrene foam sphere (inner diameter 260 mm) was painted with water-based white paint frosted spray ) And the light that can be assumed to contain no polarization component by multiple reflection inside, and then the light is incident on the sensor head unit, and the photoelectric conversion value for each wavelength is converted into the angle of the polarization filter. A calibration value recorded with the position was created in advance, and the photoelectric conversion value for each wavelength at the time of target measurement was corrected by dividing the calibration value for each rotation angle of the polarization filter.

(Calculation of conversion factor K from photoelectric conversion value by wavelength to reflection intensity)
About the corrected photoelectric conversion value for each wavelength, the standard irradiance light bulb for each wavelength of 100V500W (USHIO INC., JPD-100-500CS, verified by the Japan Electric Meters Laboratory). Measure the photoelectric conversion value of the reflected light applied to the aluminum plate (EastmanKodak: White Reflectance Coating, considered to be a perfect diffusion surface with a reflectance of 1.0) coated with barium sulfate powder using the illumination light obtained by applying the specified voltage. Then, a conversion coefficient K between the irradiance by wavelength and the photoelectric conversion value was obtained, and the photoelectric conversion value by wavelength was converted into the reflection intensity by wavelength (μW · cm ⁻² · sr ⁻¹ · nm ⁻¹ ). Thereafter, the observation results were processed based on the reflection intensity for each wavelength.

(Aquarium)
Water tanks having different soil-derived suspension concentration in water (mg / L, hereinafter sometimes referred to as SS) and bottom color were prepared as follows. As a water tank, a cylinder made of white plastic with a bottom diameter of 70cm and a height of 50cm (side wall transmittance of about 30%, bottom surface transmittance of about 20%, bottom surface reflectance of visible region 65%, near infrared region 5-60%) Using.

  Using black and tan urethane flame proof thick cloth with two layers and stitched together, the black and white sides of the light-shielding curtain are placed on the bottom of the cylinder. (Hereinafter referred to as the black bottom section; the transmissivity is negligible; the reflectance is about 2%); and the section laid with the tan surface as the upper surface (hereinafter referred to as the tan base section. The transmissivity is ignored. Possible, reflectivity is about 10-25% in the visible range, about 40% in the near infrared range), and a section where no light-shielding curtain is laid (hereinafter referred to as white bottom section. Reflectance is 65% in the visible range and 830 nm, 3 types of cylinders with different bottom colors, including 45% at 1150 nm, 28% at 1250 and 1650 nm, and 7% at 2200 nm.) Tap water into each cylinder to a depth of 45 cm. It was set as a water tank.

  A high-concentration suspension of fine-grained yellow soil collected from the surface of the paddy field in Anjo City, Aichi Prefecture, is poured into each water tank with a different color on the bottom surface, changing the injection amount, and by sufficiently stirring, SS becomes 0 mg. Five SS levels were set from / L to about 200 mg / L.

(Measurement method of reflectance and transmissivity of aquarium materials)
The reflectance is the standard diffuser plate (manufactured by Labsphere, model: SRT-99-120I, absolute) illuminated with sunlight under the same conditions as the reflection intensity of sunlight from the object using the polarized spectroradiometer. (The manufacturer's guaranteed value of reflectivity is 97% or more.)

  Similarly, the light transmittance using the polarization spectroradiometer, the reflection intensity obtained by illuminating the standard diffuser plate inclined at 45 ° with respect to the ground with sunlight, the standard diffuser plate and the It calculated | required as ratio with the reflective intensity | strength obtained when the target object was inserted perpendicularly | vertically between polarization | polarized-light spectroradiometers.

(Measurement of photoelectric conversion maximum value (Max) and photoelectric conversion minimum value (Min))
Installed aquariums with different SS levels on the open lawn in the Agricultural Environment Technology Laboratory in Tsukuba City, Ibaraki Prefecture, and used the polarized spectroradiometer to reflect the reflectance of each reflected light from each aquarium. And the polarization degree according to wavelength was measured. The measurement wavelength is center band for visible region, 3 bands of 490nm, 560nm and 660nm, center wavelength for near infrared region, 3 bands for 830nm, 1150nm and 1250nm, 2 bands for short wavelength infrared region of 1650nm and 2200nm, total 8 A band.

  The polarized spectroradiometer is installed on a tripod having a height of about 1 m, the observation field of view is set to 10 °, and the observation field of view is substantially at the center of the target water surface so that it does not reach the aquarium wall or outside. The separation distance from the aquarium water surface was adjusted in the range of 0.5m to 1.5m. The polarization spectroradiometer was provided with a sight for confirming the measurement position and range.

  The measurement was conducted on July 28. The weather was clear and windless. The experiment was conducted from 10 am to 12.30 am, during which the solar zenith angle decreased from 28 ° to 17 ° and then increased again to 19 °. The estimated value by wavelength of the amount of sunlight at the top of the atmosphere immediately above the measurement point at the time of measurement was a value obtained by multiplying the solar constant by the cosine of the solar zenith angle calculated from the latitude and longitude of the measurement point and the measurement date / time.

  The solar direction at the time of measurement was about 120 ° at the start of measurement and about 210 ° at the end of measurement when true north was 0 °, east was 90 °, and south was 180 °. The reflectance by wavelength and the degree of polarization by wavelength are measured at about 180 ° (solar facing), about 90 ° (lateral direction) and about 60 ° (diagonally behind) based on the solar direction at the time of measurement. Three angles were set as observation zenith angles, and five angles of 0, 15, 30, 45, and 60 ° were set in each of the observation directions. From the above, measurements were made for 45 SS levels of 5 levels, bottom color x 3, observation direction x 3, and observation zenith angle x 5.

  Reflected light from the object illuminated with sunlight is incident on a polarization spectroradiometer, and the reflected light passes through the bandpass filter and is split into the wavelength band. When the reflected light by wavelength passes through a polarizing filter that rotates around the optical axis, the passing light increases or decreases in accordance with the ratio of the polarization reflection intensity in the reflected light by wavelength, and accordingly, the polarization spectroradiometer The photoelectric conversion value for each wavelength also increases or decreases. Therefore, with the rotation of the polarizing filter, the maximum photoelectric conversion value (Max) and the minimum photoelectric conversion value (Min) were detected from the photoelectric conversion values obtained at every rotation angle of 1 °. This procedure is described in the literature (R. Ghosh, VN Sridhar, H. Venkatesh, AN Mehta, and KI Patel, “Linear polarization measurements of a wheat canopy”, International Journal of Remote Sensing 14, p2501-2508, 1993.) Based on.

(Calculation of polarization degree by wavelength and reflectance by non-polarization wavelength)
The polarization degree P for each wavelength was calculated from the photoelectric conversion maximum value (Max) and the photoelectric conversion minimum value (Min) by the following formula 1. The degree of polarization P for each wavelength is the ratio of the polarization reflection intensity for each wavelength in the reflected light from the object.

  P = (Max−Min) / (Max + Min) (Formula 1)

The reflectance R (%) for each wavelength was calculated by the following formula 2 using the photoelectric conversion maximum value (Max), the photoelectric conversion minimum value (Min), the conversion coefficient K, and the estimated value S for each wavelength of sunlight. .
R = K × (Max + Min) / (2 × S) × 100 (Formula 2)

  Further, the reflectance NPR (%) for each non-polarization wavelength was calculated by the following formula 3.

  NPR = R × (1-P) (Formula 3)

  Regarding the measurement results of the reflectance by wavelength and the reflectance by wavelength of non-polarized light obtained as described above, the relationship between the color of the bottom surface and SS was examined. The measured value of SS was obtained by collecting 500 ml of sample water, suction filtering with a filter (Membrane filter 0.45 μm HV, Cat. No. HVLP04700, manufactured by Millipore), and drying the captured soil particles with a filter at 105 ° C. Then, weigh and subtract the dry weight of the pre-filtered filter weighed in advance to determine the mass of the soil particles. Divide this by the volume of the sample water to obtain the number of soil mg per liter as SS (mg / L ) Was calculated.

(Result 1: SS estimation multiple regression analysis with independent polarization as a function of reflectance, observation angle, and bottom color)
Based on the 8 band measurement values obtained above, the 7-band non-polarized wave reflectance excluding 2200 nm, the solar relative observation azimuth angle, the observation zenith angle and the bottom color dummy variable are independent variables, and SS is the dependent variable. A regression analysis was performed. Note that the base color dummy variable has three base colors in this embodiment, so two variables d1 and d2 are introduced, d1 = 1, d2 = 0 for black, d1 = 0 for white , D2 = 1, and in the case of tan, d1 = d2 = -1. The results of the multiple regression analysis are shown in Table 1.

To summarize the regression analysis in Table 1, the coefficient of determination (R ² ) = 0.843, the standard deviation of error (RMSE) = 26.56 mg / L, the number of observations (N) = 213, and the observed zenith angle and solar relative observation azimuth None of the terms were significant. Therefore, it became clear that the terms of observation direction and observation zenith angle can be removed from the explanatory variables of the model by using the reflectance for each non-polarized wavelength.

(Result 2: SS estimation model consisting of 3-band non-polarized wavelength reflectance and bottom color dummy variable only)
From the analysis of the result 1, the values at the five observation zenith angles at which the reflectance by wavelength of unpolarized light was measured were averaged and summarized into 45 observation data. Next, as a result of narrowing the number of bands used using the stepwise multiple regression method such as variable increase / decrease method, by using the reflectance by wavelength of non-polarized light of 3 bands of 560 nm, 830 nm, 1650 nm and the bottom color dummy variable, A model shown in Equation 5 was obtained. The results are shown in Figure 1.

  SS = -47.25 + D−95.34 × NPR1650 + 85.32 × NPR830−4.53 × NPR560 (Formula 5)

In Equation 5, NPR1650, NPR830, and NPR560 are the non-polarized wavelength reflectances of three bands of 1650 nm, 830 nm, and 560 nm, respectively, D is the base color coefficient, black is D = 59.83, and the base color is white. , D = -80.56, and D = 20.72 when the base color is tan. The relationship between the estimated value and the actually measured value according to this equation was expressed by the coefficient of determination (R ² ) = 0.95, the standard deviation of error (RMSE) = 15.2 mg / L, and the number of observations (N) = 45.

  Therefore, Equation 5 using the non-polarized wavelength reflectance measured in the above three bands estimates the SS without considering the effect of the observation direction by adjusting the base color coefficient according to the observation target. It is effective for.

  By using the method of the present invention, the concentration of the soil-derived suspension in land water can be measured in a non-contact manner.

SS estimation by multiple regression model using non-polarized wavelength-dependent reflectance 3 bands

Claims (4)

From the reflected light of the sunlight inland water surface, the wavelength bands included in the short-wavelength infrared, unpolarized light reflectance for the wavelength band included wavelength band included in the near infrared region, and in the visible range, of the wavelength band Method for estimating the concentration of suspended matter derived from soil in inland water by calculating from the polarized light reflection intensity of the reflected light in the wavelength band detected when the reflected light passes through a polarizing filter that rotates around the optical axis .   The soil-derived suspension according to claim 1, wherein the short-wavelength infrared region is in a wavelength range of 1500 to 2300 nm, the near-infrared region is in a wavelength range of 750 to 1300 nm, and the visible region is in a wavelength range of 450 to 700 nm. Concentration estimation method. The arithmetic processing is
SS = A + D + α × NPRs + β × NPRn + γ × NPRv
The estimation method of the soil-derived suspension concentration according to claim 1 or claim 2.
In the above calculation processing, SS is the concentration of suspended soil derived from land water, D is a correction value due to the difference in bottom color, NPRs is the non-polarized reflectance of one wavelength band s included in the short wavelength infrared region, and NPRn is Non-polarized reflectance of one wavelength band n included in the near infrared region, NPRv is an unpolarized reflectance of one wavelength band v included in the visible region, and α, β, γ, and A are preset constants. Each is shown. Aperture unit for focusing light to be measured within a certain visual field range, a filter for selecting a wavelength, a polarizing filter, a polarizing filter rotating mechanism, a photoelectric conversion sensor for photoelectrically converting reflected light, and a photoelectrically converted current A device for measuring the polarization reflection intensity of wavelength band reflected light in the method for estimating the concentration of suspended matter derived from soil according to claim 1, further comprising: a voltage output device for extracting the output voltage after synchronous detection and smoothing after amplification.