Field Reflectance Measurements Research Articles

Remote Sensing of Redberry Juniper in the Texas Rolling Plains

Redberry juniper (Juniperus pinchotii Sudw.) is a noxious shrub or small tree that invades rangelands in northwest Texas. Field reflectance measurements showed that redberry juniper had lower visible and higher near-infrared (NIR) reflectance than associated species and mixtures of species in February. The low visible reflectance of redberry juniper was due to its darker green foliage than associated species, whereas its high NIR reflectance was attributed to its greater vegetative density than associated vegetation. Redberry juniper had a distinct reddish-brown image tonal response on color-infrared aerial photographs obtained in February. Computer analysis of a color-infrared photographic transparency showed that redberry juniper infestations could be quantified. An accuracy assessment performed on the classified image had a user's accuracy of 100% and a producer's accuracy of 94% for redberry juniper. DOI:10.2458/azu_jrm_v54i3_everitt

Remote sensing of redberry juniper in the Texas rolling plains

Redberry juniper (Juniperus pinchotii Sudw.) is a noxious shrub or small tree that invades rangelands in northwest Texas. Field reflectance measurements showed that redberry juniper had lower visible and higher near-infrared (NIR) reflectance than associated species and mixtures of species in February. The low visible reflectance of redberry juniper was due to its darker green foliage than associated species, whereas its high NIR reflectance was attributed to its greater vegetative density than associated vegetation. Redberry juniper had a distinct reddish-brown image tonal response on color-infrared aerial photographs obtained in February. Computer analysis of a color-infrared photographic transparency showed that redberry juniper infestations could be quantified. An accuracy assessment performed on the classified image had a user's accuracy of 100% and a producer's accuracy of 94% for redberry juniper.

Monitoring rice reflectance at field level for estimating biomass and LAI

The present study aims to monitor rice crop status during the growing season by estimating its aboveground biomass and leaf area index (LAI) from field reflectance measurements taken with a hand-held radiometer. First, vegetation indices (PVI, NDVI, WDVI, PVI) were calculated from rice crop reflectance. The fraction of intercepted photosynthetically active radiation (PAR), f PAR, is calculated based on a physical reflectance model from the vegetation indices. WDVI and PVI, indices that correct for soil reflectance, show a more linear and less-scattered relation than NDVI and RVI. The NDVI relationship with f PAR, gave a good prediction during the vegetative stage but saturated at f PAR ≥ 0.4. The soil reflectance needed for PVI and WDVI could be easily standardized for continuously flooded fields (10.2% = red reflectance; 7.0% = near-infrared reflectance). Two procedures are discussed: (a) Estimation of leaf area index (LAI) and (b) estimation of biomass. The first links f PAR with LAI by means of the extinction coefficient ( K) which varies during the growing season. The dynamic behaviour of K is analyzed and calibrated for different development stages. LAI of the rice crop was estimated directly from the calculated f PAR once K was established. This procedure accounted for only 67% of the variance in LAI. The second uses the Monteith model that requires one crop specific parameter, the conversion factor (α) for intercepted PAR into dry matter. Values of α for direct-seeded paddy rice were 2.25 g MJ −1 for total aboveground biomass over the growing period, but were smaller towards the end. Estimation of biomass from remote sensing data appears to be more reliable than estimation of LAI.

Crop residue estimation using multiband reflectance

Soil erosion is very dependent on the protection effect of residue cover. We investigated how remote sensing could be used to estimate the crop residue cover fraction. Laboratory reflectance meas-terements performed over corn residues in various moisture levels show that, for any two wave bands, a residue line similar to that for bare soils can be defined. However, it depends on the aging state of the residues, particularly in the near-infrared and middle-infrared domains. Field reflectance measurements perfornaed over a range of crop cover fraction and soil backgrounds show that a linear mixing inodel accurately describes the reflectance of the soil residue complex. An algorithm called CRIM (crop residue index multiband) is proposed. It is based on the linear mixing model of the soil-residue complex and on. soil and residue lines. It can be used with any set of wave bands. Comparison with field -measurements performed in Thematic Mapper (TM) bands TM2, TM3, TM3, TM4, and TM5 shows reasonably good estimates of the crop residue cover fraction compared with previously proposed indices (Normalized Difference Index and Soil Adjusted Corn Residue Index). The CRIM is a linear algorithm that is resistant to any linear transforms of the reflectance. It can therefore be used with. ground-level reflectance, satellite radiance, or digital counts and should not strongly depend on. view or sun geometry or on the spatial scale. The conditions for the CRIM computation and limits of its use are discussed.

High-spectral resolution field and laboratory optical reflectance measurements of red spruce and eastern hemlock needles and branches

Branch samples were collected from canopies of red spruce ( Picea rubens) and eastern hemlock ( Tsuga canadensis) on 11 and 12 September 1989 and 5 and 6 September 1990, and spectrally characterized by needle age class (first- and second-year). Needles from the branch samples were analyzed for complete optical properties (0.4–1.0 μm), pigment content (total chlorophylls), and anatomical (cellular) condition. Spectral differences between spruce and hemlock first- and second-year needles include differences in green peak reflectance features, red edge parameters, and amplitude features of the NIR plateau. Second-year needles of both species exhibit increased absorptance in the NIR when compared with first-year needles. Chlorophyll concentrations, as determined using both quantitative and empirical techniques [ratio analysis of reflectance spectra (RARS)] were greater in second-year needles in both species and highest overall in second-year hemlock needles. Substantial anatomical differences are seen between needles of the two species, as well as between age classes of the same species. Relative differences among total area occupied by cells, intercellular void space, and total needle volume likely contribute to the differences in the NIR response patterns observed. The reflectance spectra (0.4–2.5 μm) measured for single age-class branch segments of both species are similar in shape to reflectance spectra measured at the needle level. However, differences in the magnitude of reflectance are seen when branch and needle data are compared.

Remote sensing of soil color: Principles and applications

Abstract Color is an important parameter in Soil Science, widely used since the very first soil investigations. Color visually estimated in the field with the Munsell Atlas is used for soil identification and classification, for estimating various soil properties and in soil survey and mapping. The difficulties encountered in the first attempts in establishing empirical relationships between surface color and remote sensing data led to use a more scientific approach. In this paper the significance of Munsell soil color versus spectral signature is investigated and discussed with the help of colorimetric concepts and techniques. Different experiments based on field and laboratory spectral reflectance measurements are summarized. Among the results, the transformation of Munsell data into Red, Green, Blue color coordinates (R,G,B) is the basis for a pragmatic, physically based model relating surface color to satellite data in visible bands. Applications of this model to satellite imagery for soil mapping, soil mineralogy studies and arid lands vegetation monitoring are presented, with emphasis on TM data. In conclusion, the promises of remote sensing of surface color for environmental changes monitoring are discussed in anticipation of the new instruments that will fly on polar orbiting platforms at the end of this decade.

Factors causing variation in reflectance measurements from Bracken in Eastern Australia

Bracken covers large areas of cleared land in eastern Australia and is a costly problem in the management of these lands. Field reflectance measurements of patches of bracken are statistically different from those of other vegetation frequently associated with bracken in eastern Australia. There is, however, a lot of variation in reflectance measurements for any vegetation type, and some patches of vegetation may be misidentified on the basis of their reflectance characteristics. This paper examines variation in reflectance associated with shadowing, zenith angle, and percent cover of bracken. Up to 70% of the variation in groundbased measurements of reflectance from bracken and associated vegetation was accounted for by variation in shadowing and zenith angle. Variation in amount of shadowing is partially related to the age of bracken fronds, probably because the structure of the bracken canopy changes as fronds senesce. For patches of bracken mixed with other vegetation, there were gradual changes in reflectance correlated with change in amount of bracken cover. Therefore, it may be possible to estimate the percent bracken cover in a stand based on reflectance measurements, rather than simply classifying stands as “bracken” or “nonbracken.” Such estimates of percent cover may be particularly useful in Australia, where bracken often occurs mixed with other vegetation types rather than in large homogeneous stands. These results are encouraging for the prospect of using remote sensing to look at spread of bracken and/or to estimate biomass for stands of particular frond ages. Correlations between reflectance and proportion of dead bracken and between amount of bracken and reflectance are potentially useful for examining short-and long-term effects of herbicide treatments or other management strategies on large bracken stands.

The feasibility of monitoring wilderness lake chemistry with remote sensing methods

Wilderness lakes in the U.S. are at risk to the effects of atmospheric pollution. The results of the EPA Western Lake Survey, for example, show that many of the lakes in the West have very low concentrations of dissolved constituents, and are sensitive to acidification. Lakes located in wilderness areas were found to be particularly sensitive: median ANC values of 91.4 μeq/1 compared to 282.7 μeq/1 in non-wilderness areas. Sampling remote wilderness lakes is difficult and expensive. For example, the cost of sampling each lake in the National surface water survey was about $ 4000 per lake. Remote sensing methods may provide an alternative means of monitoring the chemistry of wilderness lakes. This study defines the feasibility of the remote monitoring of lake chemistry in the Adirondack Park region of New York by comparing measured lake chemistry with field reflectance measurements, remote reflectance measurements from aircraft and satellite and laboratory and airborne laser fluorosensor data.Water samples collected from a representative population of Adirondack Park lakes were analyzed for pH, acid neutralizing capacity (ANC), dissolved organic carbon (DOC), total reactive aluminum and plant pigment concentration concurrent with the collection of remote sensing data. Lake parameters estimated from field reflectance measurements which directly effect lake optical properties, total pigments, DOC, and turbidity were estimated with less error and more precision than those parameters not directly related to lake optics, pH, ANC and A1 concentration. Results from airborne MSS and Landsat MSS data produced lower R(2) values than estimates using field reflectance.The concentration of DOC can be estimated remotely by using laser fluorosensing. The spectral-fluorescence properties of DOC were correlated with the pH and aluminum concentrations in the lake water. Remotely measurable DOC fluorescence spectra were used to estimate DOC, H, and A1 concentrations.Of the methods investigated, laser fluorosensing shows the most promise for the remote prediction of lake DOC, hydrogen ion and aluminum concentration in Adirondack lakes while reflectance measures may be used to estimate lake chlorophyll, DOC and transparency. Given the large number of wilderness lakes and the difficulty of sampling in remote wilderness, remote sensing methods may provide an alternative, while less precise, method of monitoring lake chemistry. The applicability of these findings to lakes in other wilderness areas is unknown. Similar studies of the feasibility of monitoring Western wilderness lake chemistry with remote sensing methods are being initiated.

Discrimination of iron alteration minerals in visible and near‐infrared reflectance data

Field and laboratory studies were conducted to relate the mineralogy of rock and soil samples containing hematite (α‐Fe2O3), goethite (α‐FeOOH), and jarosite (KFe3(SO4)2(OH)6) to their reflectance properties in visible and near‐infrared wavelengths. Field reflectance measurements of regolith containing one or more of these minerals were made in the Goldfield, Nevada, mining district with a four‐channel radiometer. The mineralogy and reflectance properties of regolith samples collected from these field measurement sites were then determined by laboratory studies. Mossbauer spectroscopy was used to identify the iron minerals present and their relative proportions. The reflectance spectra of samples containing hematite and goethite were characterized by a reflectance minimum near 900 nm. The location of this minimum could be accounted for by the relative proportions of hematite and goethite present in the samples. Under certain conditions it may be feasible to estimate the relative proportions of hematite and goethite on the earth's surface by determining the location of this near‐infrared reflectance minimum in high spectral resolution remote sensing data. In addition, the field measurements suggest that the ratio of band passes at 852 and 982 nm may be used to distinguish qualitatively among hematite, goethite, and jarosite. Thus maps of the distribution of these minerals may be derived from scanner images acquired in those band passes.

A test of the Suits vegetative-canopy reflectance model with LARS soybean-canopy reflectance data

The Suits vegetative-canopy reflectance model is tested with an extensive set of field reflectance measurements made by the Laboratory for Application of Remote Sensing (LARS) for soybean canopies. The model is tested for the full hemisphere of observer directions as well as the nadir direction. The results show moderate agreement for the visible channels of the Landsat MSS and poor agreement in the near-infrared channel of Landsat MSS. An analysis of errors is given.

Reflectance of a vegetation canopy using the Adding method

The Adding method has been used to calculate vegetation canopy bidirectional reflectance. Appropriate reflection and transmission matrices for canopy and underlying soil layers are developed in terms of canopy geometry and basic optical properties. The model is illustrated by comparisons with field reflectance measurements for blue grama (Bouteloua gracilis) and soybeans (Calahan 9250). Calculations using the Suits model are also included for comparison.

Verification of a reflectance model for mature corn with applications to corn blight detection

A new model for calculating the directional spectral reflectance of vegetative canopies was verified for the case of mature corn by field reflectance measurements on two different mature corn fields at each of two viewing angles and over the visible and near infrared spectral bands. The application of the model to hypothetical conditions indicates that moderately severe Southern Corn Leaf Blight should be marginally detectable by aerial photography under certain special circumstances and that for more severe blight, the blight condition should be distinguishable from moisture stress. The application of the model to hypothetical fields of young corn indicates that the blight condition should be easily detected under a wide range of circumstances.