VOLUME 2 · LIGHT ENVIRONMENTS AND CANOPY ECOLOGY
冠层透光、消光与辐射截获
Canopy Light Transmission, Extinction and Radiation Interception
中英文研究初稿 R1 | 证据核查日期:2026年11月9日
Bilingual research draft R1 | Evidence review: 9 November 2026
本章把透光、截获与吸收分别定义,连接叶面积、空间结构和辐射收支,并建立可以重新计算的采样与数据规则。 它不是用一个消光系数解释全部甘蔗生理过程,而是确定每个参数在哪些观测条件下成立。
This chapter distinguishes transmission, interception and absorption, links leaf area and architecture to the radiation budget, and establishes reproducible sampling and data rules. It does not use one extinction coefficient to explain all cane physiology; it identifies the observational conditions under which each parameter is meaningful.
本章阅读路径 / Reading path
01 研究对象与核心问题 | Objects and key questions
02 结构基础与过程机制 | Structural basis and process mechanisms
03 生育阶段与环境响应 | Developmental stages and environmental responses
04 品种及生态区差异 | Genotypic and ecological differences
05 原始试验与相反证据 | Original experiments and conflicting evidence
06 观测指标与测量方法 | Variables and measurement methods
07 真实图像与机制图解 | Authentic images and interpretation
08 适用边界与研究缺口 | Applicability and research gaps
状态:初稿与首轮来源核查;尚未通过独立专家评审或全书统一审核。 Status: draft and initial source review; independent expert and whole-book approval pending.
017.01 研究对象与核心问题
Objects and key questions
1.1 从叶面积转向光的去向
From leaf area to the fate of light
第016章回答冠层有多少叶面积,本章进一步问:入射到冠层上边界的光,有多少到达下边界,有多少向上离开,又有多少留在冠层中被吸收? 同样的叶面积可以因叶角、叶片聚集、行间空隙和日照方向不同而形成不同光分布。 研究对象因此不是一个孤立的“透光百分数”,而是具有明确空间边界、波段和时间尺度的辐射收支。 [R04] [R08]
Chapter 016 quantified leaf area. This chapter asks how incoming light is transmitted through, returned above, or absorbed within a canopy. Equal leaf area does not specify the spatial distribution of light: leaf orientation, aggregation, row gaps and illumination direction also matter. The object of study is therefore a radiation budget with explicit spatial boundaries, spectral range and time support, rather than an isolated transmission percentage. [R04] [R08]
1.2 三个不能互换的量
Three quantities that are not interchangeable
透光率描述向下通量在两个高度之间的比例;几何截获描述光线是否遇到植物表面;吸收则描述光能或光子在控制体内的净损失。 实际田间常用“1减透光率”估计截获比例,但叶片透射、多次散射、反射和地表返光会使它偏离严格的几何定义。 本章把这一常用量标为表观向下截获比例 fI*,不自动改称吸收比例 fA。 [R08]
Transmittance compares downward fluxes at two heights. Geometric interception concerns whether a ray encounters a plant surface. Absorption is the net loss of radiant energy or photons inside a control volume. The common field approximation, one minus transmittance, can diverge from geometric interception because of leaf transmission, scattering, reflection and ground return. Here it is labelled the apparent downward interception fraction, fI*, and is not automatically renamed the absorbed fraction, fA. [R08]
1.3 把比较问题写成可检验问题
Specify a testable comparison
比较行距、品种或宿根年限时,应先确定要解释的是瞬时冠层下方光量、全天累计截获,还是最终糖产量。 若目标是光环境,需要同步上方基准与行内、行间观测;若目标是生产效应,还需独立小区、完整生育期和实收产量。 瞬时截光差异只能支持相应时间与空间内的判断,不能越过生长、呼吸、分配和成熟过程直接推成增糖结论。 [R01] [R04] [R06]
Before comparing spacing, cultivars or ratoon ages, identify the response: instantaneous below-canopy light, daily intercepted radiation, or harvested sugar yield. Light-environment comparisons need a synchronous reference and representative within- and between-row observations. Production claims additionally need independent plots, full-season follow-up and harvested yield. An instantaneous difference cannot bypass growth, respiration, allocation and ripening to establish a sugar-yield benefit. [R01] [R04] [R06]
1.4 控制体与研究层级
Control volume and study hierarchy
上边界宜位于冠层之上且不受邻行遮挡,下边界的位置必须写明:近地面、活叶最低层或其他固定高度。 小区侧边存在明显侧向光流时,单纯上下通量不能构成完整收支。 单叶、整茎、蔗蔸、横断面与整个地块属于不同层级;将某个行间亮点视为全田透光率,是空间代表性错误,而不是增加小数位就能解决的误差。 [R08] [M02]
The upper boundary should be above the canopy and free from neighbouring obstruction; the lower boundary must be specified, such as near the soil or below the lowest green leaf. Strong lateral flux across plot edges invalidates a purely vertical budget. Leaves, stalks, stools, transects and fields are different sampling levels. Treating a bright inter-row point as the field mean is a representativeness error that greater numerical precision cannot correct. [R08] [M02]
表017-01 量的定义与分母
Table 017-01. Quantities and denominators
指标 / Metric | 定义、单位与边界 / Definition, units and boundaries |
|---|---|
τ 透光率 / Transmittance | D_b/D_t;无量纲;同波段、同时段。 Dimensionless; matched spectral band and time. |
fI* 表观截获比例 / Apparent interception | 1−τ;描述向下通量减少,不等同于吸收。 Downward-flux removal, not necessarily absorption. |
fA 吸收比例 / Absorbed fraction | (D_t−U_t−D_b+U_b)/D_t;忽略净侧向通量。 Four-flux balance with negligible net lateral flux. |
k_eff 有效消光系数 / Effective extinction | −ln(τ)/L;需注明 L 与辐射口径。 Specify the area and radiation definitions. |
累计光量 / Radiation integral | 光子量 mol m⁻²;能量 MJ m⁻²,分别存储。 Photon and energy integrals stored separately. |
017.02 结构基础与过程机制
Structural basis and process mechanisms
2.1 四通量收支与符号
The four-flux budget and notation
令 D_t、U_t 为冠层上方的向下、向上光合有效辐射通量,D_b、U_b 为下边界的相应通量。 四者必须使用相同波段、单位与时间窗口。 在净侧向通量可忽略、无显著辐射储存的条件下,冠层吸收量由进入减离开得到。 下方反射上来的光进入冠层,所以 U_b 在吸收量公式中为正号。 这里的吸收包括绿色叶片、枯叶和茎等,不能不加区分地归入绿色叶片光合作用。 [R08]
Let D_t and U_t denote downward and upward PAR above the canopy, and D_b and U_b the corresponding lower-boundary fluxes. All four require the same waveband, units and time interval. With negligible net lateral flux and radiative storage, absorption equals inflow minus outflow. Upward radiation from below enters the canopy, explaining the positive sign of U_b. The resulting absorption includes green leaves, dead leaves and stems; it is not automatically green-leaf photosynthetic absorption. [R08]
τ = D_b / D_t; fI* = 1 − τ
APARcanopy = D_t − U_t − D_b + U_b; fA = APARcanopy / D_t
2.2 指数消光模型及其适用前提
The exponential attenuation model and its assumptions
将冠层视为统计上可描述的叶面积集合时,可用 τ≈exp(−kL) 表达透光随叶面积指数 L 的下降。 这是带有几何和散射假设的有效模型,不是对每束光的完整追踪。 对实际田间观测反求的 k,应注明波段、太阳位置、上下边界、叶面积定义和拟合方法。 将总植物面积指数作为 L,得到的参数也不能无说明地与绿色叶面积指数参数互换。 [R08] [M01]
Treating canopy elements statistically leads to the effective approximation τ≈exp(−kL), where L is leaf area index. This model embeds geometric and scattering assumptions; it is not a complete ray-by-ray description. A field-estimated k needs its waveband, sun position, measurement boundaries, area definition and fitting method. A coefficient based on total plant area cannot silently be substituted for one based on green leaf area. [R08] [M01]
τ ≈ exp(−kL); k_eff = −ln(τ) / L
Applicable inversion domain: L > 0 and 0 < τ ≤ 1.
2.3 直射、散射与叶片聚集
Direct light, diffuse light and clumping
在不透光叶片、近似随机相遇的简化直射模型中,消光取决于叶面投影函数 G、聚集指标 Ω 和太阳天顶角 θ。 散射天空来自许多方向,需要按方向辐亮度及投影权重积分,不能只指定一个“散射太阳角”。 同一冠层在不同天气下的有效 k 因而可能不同。 Ω 也不能凭透光曲线任意设定;它与叶角、叶面积以及行列空隙可能存在参数混淆。 [R08] [M02]
In a simplified direct-beam model with opaque leaves and approximately random encounters, attenuation depends on the projected-area function G, clumping index Ω and solar zenith angle θ. Diffuse illumination arrives from many directions and requires angular weighting by sky radiance and projection; there is no single diffuse solar angle. Effective k can consequently change with illumination. Clumping cannot be assigned arbitrarily from one transmission curve because it may be confounded with leaf angle, area and row gaps. [R08] [M02]
k_b(θ,φ) = G(θ,φ) Ω(θ,φ) / cos(θ) [simplified geometric model]
2.4 叶片聚集与行间孔隙不是同一现象
Leaf aggregation and row gaps are different structures
叶片在蔗蔸和茎周围聚集,会改变光线遇叶的概率;连续的宽行间隙还产生方向明确的大尺度通道。 两者可能使相同 LAI 的田块出现不同的光斑分布。 在冠层上层强光叶和下层弱光叶之间重新分配光,可能改变碳同化,但还受叶片光响应和温湿条件约束。 平均透光率只能概括通量,不能独自恢复这些结构。 [R04] [R06]
Aggregation around stalks and stools changes the probability of leaf encounters. Continuous wide inter-row gaps add directional pathways at a larger scale. These structures can generate different light patterns at equal LAI. Redistributing light between strongly illuminated upper leaves and shaded lower leaves may alter assimilation, conditional on leaf responses and microclimate. Mean transmittance summarizes flux but cannot uniquely reconstruct these structures. [R04] [R06]
2.5 截光增加的边际收益
Marginal gains in interception
在固定 k 的假设模型中,每增加一单位 L 所带来的截获增量随冠层变密而下降。 所谓“达到95%截光”的 L 取决于 k,而非全球统一的最佳 LAI。 该阈值只是光学计算,不能单独决定栽培密度:叶片维护成本、病害微气候、倒伏和糖分分配还需另外评价。 反过来,较小 k 在既定 L 下并不意味着较多总截光,但可能对应不同的垂直分配。
In the fixed-k approximation, the interception gained from another unit of L declines as the canopy closes. The L required for 95% interception depends on k and is not a universal optimal LAI. This is an optical threshold, not a stand-density prescription: maintenance costs, disease microclimate, lodging and sugar allocation require separate assessment. Conversely, lower k does not mean greater total interception at fixed L, although it may imply a different vertical distribution.
dfI*/dL = k exp(−kL)
L95 = −ln(0.05) / k [model threshold, not a field recommendation]
017.03 生育阶段与环境响应
Developmental stages and environmental responses
3.1 未封行阶段的空间异质性
Spatial heterogeneity before row closure
萌芽到早期分蘖阶段,裸地、蔗行和缺株区共同决定下方通量。 此时仅在行内放置传感器,容易遗漏最亮的行间区域;仅测行间中心,又会夸大透光。 应以完整重复行距单元为采样框架,记录探头位置与对应地面面积。 宽窄行配置至少覆盖窄间隙、宽间隙及行内,不把相同观测次数误当相同面积。 [R01] [R06]
Before closure, bare ground, crop rows and missing stools jointly determine below-canopy flux. Sensors only within rows can omit bright gaps, whereas sensors only at inter-row centres can overstate transmission. Sampling should span a full repeating row-spacing unit, recording position and represented ground area. Paired rows require narrow gaps, wide gaps and within-row positions; equal numbers of readings need not represent equal areas. [R01] [R06]
3.2 快速扩展阶段的参数变化
Changing parameters during rapid expansion
展叶、分蘖成活和茎伸长可同时改变 L 与受光结构,因此跨日期的 −ln(τ) 与 L 回归同时混合了叶面积效应和结构变化。 宜保留每次观测的太阳角、绿叶比例、茎密度和阶段标签,检查单一斜率是否留下系统性残差。 若不同阶段需要不同参数,应在独立日期或地块检验,而不是仅以训练数据拟合变好作为依据。 [R02] [R05]
Leaf expansion, tiller survival and elongation can change L and architecture together. A regression of −ln(τ) against L across dates therefore combines area effects with structural change. Retain sun angle, green fraction, stalk density and stage labels and inspect systematic residuals from a single slope. Stage-dependent parameters require validation on independent dates or plots, not merely improved fit to the calibration observations. [R02] [R05]
3.3 衰老、枯叶保留与倒伏
Senescence, retained dead leaves and lodging
成熟期活叶减少,不一定立即产生相同比例的透光增加,因为枯叶、叶鞘和茎仍可遮光。 倒伏会同时改变叶面倾角、相互覆盖以及传感器上方实际植被量。 测量应分别登记绿色叶面积、总遮挡面积的代理指标和倒伏状态。 不能把光学反演得到的高面积指数直接解释为较多功能叶,更不能由其推定较高糖产量。 [M01] [R08]
Declining green leaf area need not immediately produce proportional transmission increases because dead leaves, sheaths and stalks still obstruct light. Lodging changes orientation, overlap and the vegetation actually above a sensor. Record green area, an optical obstruction proxy and lodging separately. A high optically inferred area index does not establish more functional leaf area or greater sugar yield. [M01] [R08]
3.4 日内变化及水分影响的区分
Separating diurnal geometry from water effects
同一天上午与下午即使上方 PAR 相同,太阳方位、叶片卷曲或移动仍可能改变透光。 干旱还可通过抑制扩叶与加速叶衰退改变季节性 L。 因此要区分直接几何效应、慢速生长效应和叶片光合受限。 记录风况、叶片状态和水分处理有助于解释变化,但未经控制的相关性不足以确定某个原因占比。 [R03] [M02]
Equal above-canopy PAR in morning and afternoon does not ensure equal transmission: solar azimuth, leaf rolling or motion can differ. Water deficit can also alter seasonal L through slower expansion and faster senescence. Separate immediate geometry, slower growth effects and limitations of leaf photosynthesis. Wind, leaf condition and water-treatment records aid interpretation, but uncontrolled associations cannot assign a unique causal contribution. [R03] [M02]
表017-02 生育阶段与采样重点
Table 017-02. Stages and sampling priorities
阶段 / Stage | 优先记录 / Priority records |
|---|---|
未封行 / Open rows | 完整行距单元、缺株、裸地与行内位置。 Complete row units, missing stools, bare ground and within-row positions. |
快速展叶 / Rapid expansion | 同步 LAI、叶角、茎密度与太阳位置。 Synchronous LAI, orientation, stalk density and solar position. |
封行 / Closed canopy | 下层低通量检测限、空间变异和枯叶。 Low-flux detection limits, spatial variation and dead leaves. |
成熟或倒伏 / Ripening or lodging | 绿叶比例、器官遮挡与冠层高度改变。 Green fraction, non-leaf obstruction and height changes. |
017.04 品种及生态区差异
Genotypic and ecological differences
4.1 把品种效应与观测条件拆开
Separate cultivar differences from observation conditions
比较品种的 k,应在相同波段与测量方案下,对齐物候、太阳位置和有效叶面积口径。 早熟品种在同一日历日期可能已经进入衰老,晚熟品种仍在展叶;这种比较描述的是当时的综合表现,不是“纯叶角基因效应”。 相同日龄和相同发育阶段两套比较均有价值,但回答的问题不同。 [R02] [R05]
Cultivar comparisons of k require a common waveband and protocol, with phenology, solar position and area definition aligned. On one calendar date an early cultivar may be senescing while a later cultivar is still expanding. That contrast describes integrated performance at that date, not a pure leaf-angle genetic effect. Equal-age and equal-stage comparisons are both useful but address different questions. [R02] [R05]
4.2 生态区不能用一个国家标签代替
An ecological environment is more than a country label
同一蔗区也可具有不同云量、种植季节、行向、土壤水分及宿根制度。 数据表应保留地点坐标、气候观测时段、作物起始事件和管理方式,再解释环境差异。 不能只按“中国、巴西、南非”等标签平均 k;同样不能把不同仪器波段、拟合方法和太阳角造成的差异全部归因于生态适应。 [R01] [R03] [R05] [R06]
Within one cane region, cloudiness, planting season, row orientation, water status and ratoon management may differ. Preserve coordinates, meteorological observation periods, crop starting events and management before interpreting environmental effects. Averaging k by country alone is not defensible. Differences in instrument waveband, fitting method or sun angle must not all be attributed to ecological adaptation. [R01] [R03] [R05] [R06]
4.3 跨地区模型的检验单位
Units of cross-region model validation
模型可先在一个地点拟合叶面积和光分布,再到另一个地点检验累计截获、干物质与误差结构。 保留不同季节与宿根类别,而不是把所有时间点随机混合后分成训练集和验证集;相邻日期往往共享同一小区和生长历史。 地点外检验应同时报告偏差、散布和适用范围,不能只给一个相关系数。 [R05]
A model can be calibrated for area and light distribution at one site and tested for cumulative interception, biomass and error structure elsewhere. Preserve season and ratoon categories rather than randomly dividing all timestamps into training and validation sets; adjacent dates often share a plot and growth history. External evaluation should report bias, scatter and domain of applicability, not correlation alone. [R05]
017.05 原始试验与相反证据
Original experiments and conflicting evidence
5.1 行距试验的参数不等于独立因果验证
Spacing coefficients do not establish an independently replicated effect
巴西 RB98710 研究报告单行与宽窄行的拟合 k 分别为0.48±0.03和0.51±0.02;方法显示每种配置仅一个地块。 10株生长观测和多次辐射测量是地块内或时间子样本,不能弥补处理重复不足。 本文保留原文“±”,但未确认其统计类型,不改标为标准误或标准差。 这些系数可用于理解该研究,不能单独证明某行距普遍优越。 [R01]
The RB98710 study reported fitted k of 0.48±0.03 and 0.51±0.02 for single and paired rows, but its methods describe only one plot per configuration. Plant subsamples and repeated radiation campaigns do not supply missing treatment replication. The reported ± is retained without relabelling it as SD or SE. These coefficients describe that study; they do not establish a generally superior spacing. [R01]
5.2 冠层图像参数与通量参数的区别
Image-derived and flux-derived parameters differ
福建17品种研究使用鱼眼图像描述五个阶段的冠层结构,报道叶面积与消光参数具有不同季节轨迹。 其方向性图像参数不应直接拼接到水平 PAR 通量回归中。 研究的产量由形态性状推算,不能与独立实收产量混同;较强相关性也不是冠层结构直接导致增产的充分证明。 [R02]
The Fujian study used fisheye images across five stages in 17 varieties and reported differing seasonal trajectories for area and extinction parameters. Its directional image indices should not be pooled uncritically with horizontal PAR-flux regressions. Yield was estimated from morphological traits rather than measured as independent harvested mass; strong association alone is insufficient evidence that canopy structure caused a yield increase. [R02]
5.3 灌溉对照与波段、误差层级
Irrigation contrasts, wavebands and error strata
斯里兰卡试验在裂区设计中配置灌溉主区与品种副区,以管状太阳辐射仪测上下方辐射。 其水分效应和品种效应必须使用对应误差层级。 本章已核读方法文字,但未成功取得 PDF 表图页面图像,因此不从其结果表提取消光数值;仪器光谱响应未明确时,也不把结果改称400—700 nm 的 PAR 专用参数。 [R03]
The Sri Lankan experiment used irrigation main plots and variety subplots, measuring above- and below-canopy radiation with a tube solarimeter. Water and variety contrasts require their respective error strata. Methods text was inspected, but PDF table-image verification was unsuccessful, so no extinction results are transcribed from those tables. Without an established spectral response, its coefficient is not relabelled as specific to 400–700 nm PAR. [R03]
5.4 同样叶面积仍可能产生不同空间结果
Equal area can still produce different spatial outcomes
三维甘蔗模型在相同 LAI 和材料条件下比较100 cm等行距与45/155 cm交替行距,预测不均匀配置的冠层净碳吸收约低9%。 这是模型情景输出,不是实测蔗糖减产率。 每个情景的三次虚拟冠层重复仅反映构型随机性,不能替代田间生物学重复。 它支持进一步检验空间配置,而不是宣布一个全球固定行距。 [R04]
A three-dimensional cane model compared 100 cm uniform rows with alternating 45/155 cm spacing at equal LAI and material settings, predicting about 9% lower canopy net carbon uptake under the uneven arrangement. This is a scenario prediction, not a measured sugar-yield loss. Three virtual-canopy repetitions reflect architectural randomness, not field biological replication. The result motivates testing spatial configuration rather than prescribing a universal row spacing. [R04]
5.5 覆盖度可相关,但不与截光相等
Coverage can correlate with interception without being identical
甘蔗 Canopeo 研究报告图像覆盖度与传感器截光指标的相关系数 r=0.764。 照片投影的绿色像素比例和从天空多方向进入冠层的光,几何定义并不相同。 方法与结果中的拍照、PAR观测时间表述仍需对照核实,因此本章只保留相关性结果,不将其变成可跨时段替代传感器的通用换算式。 [R07]
The cane Canopeo study reported r=0.764 between image coverage and a sensor-based interception index. Green-pixel projected cover and radiation arriving from multiple sky directions have different geometries. Timing descriptions for photographs and PAR observations still need reconciliation. The reported correlation is retained, but not turned into a universal equation allowing photographs to replace radiation sensors across times. [R07]
5.6 支持、不一致与缺口同时保留
Retain support, inconsistency and gaps together
现有研究共同提示要考虑空间结构,但并未给出完全等价的 k:仪器、面积口径、时间窗口和处理设计不同。 27品种研究的摘要提供了跨材料模型参数化线索;行内与行间分别观测的研究强调采样位置。 这些来源适合确定研究方向,尚不足以重建所有试验的独立数据。 水稻四通量方法提供物理定义,但其具体仪器表现不是甘蔗验证结果。 [R05–R08]
The studies support attention to spatial structure but do not supply interchangeable k values: instruments, area definitions, time windows and designs differ. The 27-variety abstract offers a lead for cross-genotype parameterization, while separate within- and between-row observations emphasize sampling position. These sources guide research without reconstructing all independent trial data. The rice four-flux analysis supplies physical definitions, not cane-specific instrument validation. [R05–R08]
表017-03 原始研究及证据边界
Table 017-03. Original studies and limits
来源 / Source | 能回答的问题与限制 / Contribution and limitation |
|---|---|
R01 巴西 / Brazil | 田间通量拟合;每行距仅一地块,不构成重复处理验证。 Flux fitting; one plot per spacing, not replicated treatment evidence. |
R02 中国 / China | 季节性图像结构;方向性参数与推算产量。 Seasonal image structure; directional parameters and calculated yield. |
R03 斯里兰卡 / Sri Lanka | 裂区水分试验方法;波段不明,表图未视觉核验。 Split-plot water methods; unresolved band and unverified table images. |
R04 三维模型 / 3D model | 等 LAI 空间对照;碳吸收预测,不是实测糖产量。 Equal-LAI geometry comparison; predicted carbon, not harvested sugar. |
R05 巴西 / Brazil | 多品种参数化线索;摘要层级。 Multi-variety parameterization lead; abstract-level access. |
R06 南非 / South Africa | 行内与行间区分;完整设计待取得。 Within-/between-row contrast; full design not acquired. |
R07 印度 / India | 图像覆盖与截光关联;时间一致性及独立重复待核。 Cover–interception association; timing and independence unresolved. |
R08 水稻方法 / Rice methods | 四通量定义;不是甘蔗田间验证。 Four-flux definitions; not field validation in sugarcane. |
017.06 观测指标与测量方法
Variables and measurement methods
6.1 同步基准与横断面布设
Synchronous reference and transect layout
本章建议在每个独立小区设置无遮挡上方 PAR 基准,地面或下边界探头沿完整行距单元布设,并记录每个位置的面积权重。 移动探头时保持上方连续记录;多小区应随机化测量顺序,避免处理与云影时间固定重合。 若研究吸收,还需上、下边界的向上通量;若只有两个向下探头,就如实报告透光或 fI*,不补造反射数据。 [R08] [M02]
The proposed protocol places an unobstructed above-canopy PAR reference in each independent plot and samples the lower boundary across complete row-spacing units with area weights. Keep the reference logging while moving a probe. Randomize plot order to avoid confounding treatments with cloud passage. Absorption requires upward fluxes at both boundaries; with only two downward measurements, report transmission or fI* rather than inventing reflection data. [R08] [M02]
6.2 标定、位置与低通量质量控制
Calibration, position and low-flux quality control
测量前并排比较传感器,登记零点、比例响应、水平状态、清洁情况和校准日期。 上方信号很低时,比值对噪声敏感;下方值接近检测限时,不能记为精确零。 若 D_b>D_t,不应直接压成100%:先查不同步、倾斜、遮挡、侧向入光及云边变化。 保留原始值与质量标记,再决定排除或采用专门模型。 [M02]
Before measurement, compare sensors side by side and record zero response, relative sensitivity, levelling, cleanliness and calibration date. Ratios become noise-sensitive at low reference flux; below-canopy readings near detection limits are not exact zero. When D_b>D_t, do not simply cap transmission at 100%: investigate asynchrony, tilt, obstruction, lateral light and changing clouds. Retain raw values and quality flags before exclusion or specialized modelling. [M02]
6.3 先确定空间估计量,再进行对数变换
Define the spatial estimator before taking logarithms
估计整个小区的通量透过率时,先按面积权重汇总上下方通量,再计算比例。 对每点取对数后平均,得到的是另一种统计量。 由于对数非线性,两个结果通常不同;不能把差异全部说成仪器误差。 若上方基准随位置或时间不同,还需先匹配对应参考观测,不能直接平均各点未经加权的百分数。
To estimate plot-level flux transmission, combine fluxes using represented-area weights before calculating the ratio. Averaging pointwise logarithms estimates a different quantity. The logarithm is nonlinear, so these results generally differ; the discrepancy is not simply instrument error. When reference light differs among positions or times, match the corresponding observations before aggregation rather than averaging unweighted percentages.
τplot = Σ(wᵢ D_b,ᵢ) / Σ(wᵢ D_t,ᵢ), Σwᵢ = 1
−ln[mean(τ)] ≠ mean[−ln(τ)] in general.
6.4 日积分、季节积分与分母
Daily and seasonal integration with the correct denominator
若通量单位为 μmol m⁻² s⁻¹,用每段通量乘持续秒数并乘10⁻⁶,得到 mol m⁻² 的累计光子量;W m⁻² 的时间积分则产生能量,不能用同一数值冒充光子量。 日截获比例应为累计截获量除累计入射量。 季节比较还需统一起止事件和缺测处理,不将生长季更长造成的总量差直接解释为利用效率差。
Flux in μmol m⁻² s⁻¹ multiplied by interval duration in seconds and 10⁻⁶ gives cumulative mol m⁻². Integrating W m⁻² yields energy, not interchangeable photon counts. The daily fraction is cumulative intercepted radiation divided by cumulative incident radiation. Seasonal comparisons also require aligned starting and ending events and missing-data rules; longer duration alone does not demonstrate better conversion efficiency.
IPAR* = 10⁻⁶ Σ[(D_t,j − D_b,j) Δtⱼ] [mol m⁻² for photon-flux input]
fI*,day = Σ[Qⱼ fI*,ⱼ] / ΣQⱼ [Qⱼ = incident radiation in interval j]
6.5 拟合k与参数不确定性
Fitting k and quantifying uncertainty
−ln(τ) 对 L 的斜率法要求说明截距是否固定、误差位于哪个变量以及同一小区的重复时间相关性。 L 本身有误差时,普通最小二乘可能不适合;应检查残差,并考虑误差变量模型或带小区随机效应的模型。 近似误差传播式只能用于相应独立误差假设;共享参考传感器造成的协方差应保留。
A slope fit of −ln(τ) against L must state whether the intercept is constrained, which variables contain error and how repeated times within plots are handled. Error in L can make ordinary least squares inappropriate. Inspect residuals and consider errors-in-variables or plot-level mixed models. Approximate propagation formulas require their stated independence assumptions; covariance from a shared reference sensor must be retained.
Var(k) ≈ [σ(τ)/(Lτ)]² + [ln(τ) σ(L)/L²]² [independent-error approximation]
6.6 模型检验与避免循环验证
Model testing without circular validation
以透光反演 LAI,再把同一 LAI 和同一 k 代回指数式重建透光,所得良好吻合不是独立验证。 至少需要另一种叶面积测量,或保留独立小区、年份的光通量检验。 半球图像得到的方向间隙率、地面覆盖度和上下方 PAR 通量应作为不同观测流保存,检验它们之间的关系,而不是先强行设为相等。 [M01] [M02] [R07]
Inverting LAI from transmission and inserting it back into the same equation with the same k does not provide independent validation. Use an independent area measurement or reserve plots and years for flux evaluation. Directional gap fractions from hemispherical images, projected ground cover and paired PAR fluxes should remain separate observation streams whose relationships are tested rather than imposed. [M01] [M02] [R07]
6.7 结构化记录与可复算性
Structured records and reproducibility
每条观测至少关联地点、小区、处理、传感器、位置、时间、波段、单位和质量标记;计算结果还要指向输入观测及公式版本。 模型拟合参数、发表均值、原始传感器读数和教学算例分表保存。 本章提供的发表数字不是重新取得的逐小区原始数据;空值表示未取得或不适用,不能用零补齐。
Every observation should link site, plot, treatment, sensor, position, time, waveband, units and quality flags. Derived values must reference their inputs and formula version. Store fitted parameters, published summaries, original sensor readings and teaching examples separately. Published values transcribed here are not newly acquired plot-level raw data. Null denotes unavailable or inapplicable information and is not filled with zero.
表017-04 最小可追溯观测记录
Table 017-04. Minimum traceable observation record
字段组 / Field group | 关键字段 / Key fields |
|---|---|
材料与处理 / Material and treatment | site_id, plot_id, cultivar, crop_cycle, treatment_id;小区为统计层级。 Plot identity retained as a statistical level. |
传感器与位置 / Sensor and position | sensor_id, calibration_id, waveband_nm, height_m, row_position, area_weight。 |
时间与通量 / Time and flux | timestamp, timezone, interval_s, D_top, U_top, D_bottom, U_bottom, unit。 |
面积与环境 / Area and environment | LAI_method, green_fraction, sun_angle, sky_condition, lodging, leaf_wetness。 |
质量与推导 / Quality and derivation | quality_flag, detection_limit, missing_reason, input_ids, formula_version。 |
表017-05 教学演算:均为假设值
Table 017-05. Teaching calculations: hypothetical values only
算例 / Example | 输入与结果 / Inputs and results |
|---|---|
E1 四通量 / Four fluxes | D_t=1600, D_b=400, U_t=120, U_b=40 μmol m⁻² s⁻¹。 τ=0.25;fI*=0.75;APAR=1120;fA=0.70。 |
E2 95%阈值 / 95% threshold | k=0.5 → L95≈5.99;k=0.3 → L95≈9.99。 模型阈值,不是田间推荐 / Model threshold, not a field prescription. |
E3 空间平均 / Spatial averaging | 等面积 τ=0.1 和0.8,L=3:先平均通量 k≈0.266;先逐点反演再平均 k≈0.421。 Equal-area flux averaging and local-log averaging estimate different quantities. |
E4 时间加权 / Time weighting | 入射20和5 mol m⁻²,截获比例0.5和0.9。 累计截获14.5 mol m⁻²;加权比例0.58,非算术平均0.70。 |
017.07 真实图像与机制图解
Authentic images and interpretation
7.1 背景实拍能说明什么
What a contextual photograph can establish
图017-01沿用已有甘蔗田原始照片,展示冠层、田缘和天空背景。 它没有同步辐射计、处理标签或叶面积测量,只能作为现场环境的直观参照。 摄影曝光、白平衡和云层外观均不能换算出该田的 PAR、透光率或 k。 图像保留原件哈希、摄影者、日期和许可,与实验数据分开登记。 [M03]
Figure 017-01 reuses the original field photograph to show canopy, field edge and sky context. It has no synchronous radiometer readings, treatment labels or leaf-area measurements and serves only as observational context. Exposure, white balance and visible clouds cannot determine the field PAR, transmission or k. Original-file hash, creator, date and licence are retained separately from experimental data. [M03]

网页预览经过缩放或格式转换,未裁切;原比例尺与标签保留。 / Web preview resized or re-encoded without cropping; original scales and labels retained.
图017-01 甘蔗冠层、田缘与天空背景实拍。 摄影:Kerry Raymond,2022年1月26日,澳大利亚昆士兰州Alberton,CC BY 4.0。 无同步辐射定标,不能据此计算透光率、LAI或k。 Figure 017-01. Authentic canopy, field-edge and sky context. Photo: Kerry Raymond, 26 January 2022, Alberton, Queensland, Australia; CC BY 4.0. Without synchronized radiometric calibration, this image cannot yield transmittance, LAI or k.
7.2 照片、分类图和数据图分开入库
Separate photographs, classifications and data plots
文献中的冠层照片、二值绿色覆盖图和散点回归图具有不同证据角色。 Canopeo 论文图2含有这些不同成分,不能把整版均称为“田间实拍”;图1则为示意绘图。 该候选图原件本章下载失败,已停止重试,未计入本章已取得图像。 在线能看到图注或缩略图,不等于已保存原件或完成图版权属核验。 [R07]
Canopy photographs, binary green-cover classifications and regression plots play different evidential roles. Figure 2 in the Canopeo paper combines such components and is not wholly a field photograph; Figure 1 is a drawing. The candidate original could not be downloaded in the source review, so acquisition was stopped and it is not counted as obtained media. An online caption or thumbnail is not a saved original or completed figure-rights review. [R07]
7.3 优先补采的实证图像
Priority images for future evidence collection
应优先补采同一独立小区的上方基准、完整横断面探头位置、行距尺度和同步时钟照片,再配合不同生育阶段重复拍摄。 需要标明摄影方向、探头高度、是否同日同处理,以及叶片是否人为拨开。 机制示意可以帮助解释四通量和采样权重,但必须显式标为示意,不能代替有传感器记录的处理对照。
Priority images should document the above-canopy reference, complete transect positions, spacing scale and synchronized clocks in the same independent plot, repeated across stages. Record viewing direction, sensor height, date and treatment alignment and any deliberate displacement of leaves. Mechanistic diagrams can explain four fluxes and sampling weights but must remain explicitly schematic, never substitutes for treatment comparisons with sensor records.
017.08 适用边界与研究缺口
Applicability and research gaps
8.1 能够支持与不能支持的结论
Supported interpretations and unsupported extensions
本章支持用有边界、有单位的通量描述透光,并在明确假设下估计有效消光系数;不支持规定全球统一 k、最优 LAI 或最优行距。 高截获不等于高吸收,高吸收不等于高净碳收益,高净碳收益也不等于高可回收糖。 后续第022章应在相同辐射分母下讨论利用效率,避免把不同层次的指标互相替代。 [R04] [R08]
This chapter supports bounded, unit-explicit flux measurements and effective attenuation estimates under stated assumptions. It does not establish a universal k, optimal LAI or row spacing. Greater interception is not automatically greater absorption, net carbon gain or recoverable sugar. Chapter 022 should discuss conversion efficiency with a consistent radiation denominator rather than substituting indicators across these levels. [R04] [R08]
8.2 关键缺口及可证伪研究
Key gaps and falsifiable studies
当前缺口包括独立地块重复不足、波段与测量窗口不统一、部分论文仅有摘要,以及带同步辐射记录的真实图像不足。 一个优先研究是:在重复小区中同时测四通量、直接绿叶面积与行距断面,检验使用单一 k 和分阶段 k 的预测差,并以另一个年份验证。 应预先写明何种误差降低才有实用意义,也应如实报告无改进结果。
Current gaps include insufficient independent plot replication, inconsistent wavebands and measurement windows, abstract-only access for some papers and limited images linked to synchronous flux records. A priority test would measure four fluxes, direct green area and row transects in replicated plots, compare constant and stage-dependent k, and validate in another year. Prespecify a practically meaningful reduction in prediction error and report null improvements as well.
8.3 首轮核查与出版边界
Initial review and publication boundary
本章完成的是双语研究初稿与来源核查,不等于独立专家评审或全书定稿。 第003项来源的表图视觉核验、第006项来源的完整设计、第007项来源的测量时间一致性仍需补齐;不得通过增加主张等级消除资料缺口。 下一章转入冠层上、中、下层功能叶的光环境。
This is a bilingual research draft with an initial source review, not independent expert approval or final whole-book certification. Outstanding tasks include table-image verification for R03, the full design of R06 and timing consistency in R07; assigning stronger labels cannot remove missing information. The next chapter examines functional-leaf light environments in upper, middle and lower canopy layers.
表017-06 判断性主张的审核路径
Table 017-06. Review paths for interpretive claims
拟提出主张 / Proposed claim | 必须追加的证据 / Additional evidence required |
|---|---|
“该品种 k 更低” / “This cultivar has lower k” | 相同波段、太阳角与阶段;独立小区;面积测量误差。 Matched band, angle and stage, independent plots and area uncertainty. |
“较低 k 增糖” / “Lower k increases sugar” | 截光以外的生理链及跨季实收糖量。 Mechanistic links beyond interception and multi-season harvested sugar. |
“照片替代传感器” / “Photos replace sensors” | 独立验证、偏差分析、不同光照和冠层条件。 Independent validation, agreement analysis and varied light/canopy conditions. |
“模型验证通过” / “The model is validated” | 非循环验证,保留地点、年份或小区层级。 Non-circular tests separated by site, year or plot. |
附录A 双语术语与符号
Appendix A. Bilingual terminology and symbols
表017-07 本章术语
Table 017-07. Chapter terminology
中文 | English | 符号 / Symbol |
|---|---|---|
冠层透光率 | canopy transmittance | τ |
几何截获 | geometric interception | |
表观截获比例 | apparent interception fraction | fI* |
吸收比例 | fraction absorbed | fA |
光合有效辐射 | photosynthetically active radiation | PAR |
光合光子通量密度 | photosynthetic photon flux density | PPFD |
截获光合有效辐射 | intercepted PAR | IPAR |
吸收光合有效辐射 | absorbed PAR | APAR |
向下通量 | downward flux | D |
向上通量 | upward flux | U |
净侧向通量 | net lateral flux | |
控制体 | control volume | |
叶面积指数 | leaf area index | LAI |
植物面积指数 | plant area index | PAI |
有效消光系数 | effective extinction coefficient | k_eff |
间隙率 | gap fraction | P_gap |
叶面投影函数 | leaf projection function | G |
聚集指数 | clumping index | Ω |
太阳天顶角 | solar zenith angle | θ |
方位角 | azimuth angle | φ |
直射光 | direct-beam radiation | |
散射光 | diffuse radiation | |
反射 | reflection | |
多次散射 | multiple scattering | |
行内 | within-row | |
行间 | between-row | |
面积权重 | area weight | w |
辐射加权 | radiation weighting | |
半球摄影 | hemispherical photography | DHP |
检测限 | detection limit | |
删失观测 | censored observation | |
实验单位 | experimental unit | |
子样本 | subsample | |
误差变量模型 | errors-in-variables model | |
独立验证 | independent validation | |
循环验证 | circular validation |
附录B 原始来源与访问边界
Appendix B. Original sources and access boundaries
文献中的试验、模型和方法支持分别记录;同一研究的论文、图片与数据库登记不重复计为独立验证。 以下链接是原始来源入口,不表示本章已下载可再分发全文。 Experiments, models and method support are distinguished. Papers, figures and database records from one study are not independent confirmations. Links below identify sources, not newly downloaded redistributable full texts.
[R01] Ferreira Junior, R.A.; de Souza, J.L.; Lyra, G.B.; Escobedo, J.F.; Santos, M.V.C. (2015). Energy conversion efficiency in sugarcane under two row spacings in northeast of Brazil
巴西东北部两种行距配置下甘蔗的能量转化效率
DOI: 10.1590/1807-1929/agriambi.v19n8p741-747
https://www.scielo.br/j/rbeaa/a/KKCQy59XVwyXbMb6KcgFV8K/?lang=en
访问层级 / Access: 网页全文:方法与结果 / Full HTML: methods and results
[R02] Luo, J.; Que, Y.; Zhang, H.; Xu, L. (2013). Seasonal Variation of the Canopy Structure Parameters and Its Correlation with Yield-Related Traits in Sugarcane
甘蔗冠层结构参数的季节变化及其与产量相关性状的关系
DOI: 10.1155/2013/801486
https://onlinelibrary.wiley.com/doi/10.1155/2013/801486
访问层级 / Access: 网页全文:方法与结果 / Full HTML: methods and results
[R03] De Silva, A.L.C.; De Costa, W.A.J.M. (2012). Growth and Radiation Use Efficiency of Sugarcane Under Irrigated and Rain-fed Conditions in Sri Lanka
斯里兰卡灌溉与雨养条件下甘蔗生长及辐射利用效率
DOI: 10.1007/s12355-012-0148-y
访问层级 / Access: PDF文字已核读;表图页面未核验 / PDF text inspected; table/figure images unverified
[R04] Wang, Y.; Song, Q.; Jaiswal, D.; de Souza, A.P.; Long, S.P.; Zhu, X.-G. (2017). Development of a Three-Dimensional Ray-Tracing Model of Sugarcane Canopy Photosynthesis and Its Application in Assessing Impacts of Varied Row Spacing
甘蔗冠层光合三维光线追踪模型及其行距效应应用
DOI: 10.1007/s12155-017-9823-x
https://link.springer.com/article/10.1007/s12155-017-9823-x
访问层级 / Access: 网页全文:方法与结果 / Full HTML: methods and results
[R05] Dias, H.B.; Inman-Bamber, G.; Everingham, Y.; Sentelhas, P.C.; Bermejo, R.; Christodoulou, D. (2020). Traits for canopy development and light interception by twenty-seven Brazilian sugarcane varieties
27个巴西甘蔗品种的冠层发育与截光性状
DOI: 10.1016/j.fcr.2020.107716
https://researchonline.jcu.edu.au/63033/
访问层级 / Access: 作者机构摘要 / Author-institution abstract
[R06] Singels, A.; Smit, M.A. (2009). Sugarcane response to row spacing-induced competition for light
甘蔗对行距引起的光竞争的响应
DOI: 10.1016/j.fcr.2009.04.015
https://www.sciencedirect.com/science/article/abs/pii/S0378429009001063
访问层级 / Access: 出版商索引摘要与方法片段 / Publisher-indexed abstract and methods excerpt
[R07] Kumar, R.A. et al. (2023). Rapid and Non-Destructive Methodology for Measuring Canopy Coverage at an Early Stage and Its Correlation with Physiological and Morphological Traits and Yield in Sugarcane
甘蔗早期冠层覆盖度快速无损测量及其与生理形态和产量的关系
DOI: 10.3390/agriculture13081481
https://www.mdpi.com/2077-0472/13/8/1481
访问层级 / Access: 出版商索引方法与结果片段 / Publisher-indexed methods/results excerpts
[R08] Li, W.; Fang, H.; Wei, S.; Weiss, M.; Baret, F. (2021). Critical analysis of methods to estimate the fraction of absorbed or intercepted photosynthetically active radiation from ground measurements: Application to rice crops
地面观测吸收或截获光合有效辐射比例的方法辨析:以水稻为例
DOI: 10.1016/j.agrformet.2020.108273
https://iarrp.caas.cn/en/docs/2022-07/8f9cfe5f4d0c4b548d34fcc691398d3e.pdf
访问层级 / Access: PDF文字及公式页已核验 / PDF text and equation page verified
[M01] LI-COR (n.d.). LAI-2200C: Theory of Operation
LAI-2200C工作原理
https://www.licor.com/support/LAI-2200C/topics/theory-of-operation.html
访问层级 / Access: 官方方法网页 / Official method webpage
[M02] METER Group (n.d.). The researcher's complete guide to Leaf Area Index (LAI)
叶面积指数研究指南
https://metergroup.com/education-guides/the-researchers-complete-guide-to-leaf-area-index-lai/
访问层级 / Access: 官方方法网页 / Official method webpage
[M03] Kerry Raymond (2022). Sugarcane fields, looking south down Cane Farm Road, Alberton, 2022 01
Alberton甘蔗田与田缘实景
访问层级 / Access: 已有图像元数据与本章像素/哈希核验 / Inherited media provenance; current pixel/hash verification
附录C 证据记录的解释
Appendix C. Interpreting the evidence records
配套JSON逐条保存来源、主张、数值、教学计算、图像与缺口。 全文可读不等于强证据;研究设计与适用性必须单独评价。 数十个数据字段也不意味着相同数量的独立观测。 本文没有取得逐小区原始数据,因此不存在基于新原始数据的再分析结论。
The companion JSON stores sources, claims, numeric records, teaching calculations, media and gaps separately. Full-text access is not evidence strength; design and applicability are assessed separately. The number of fields or records is not the number of independent observations. No plot-level raw dataset was acquired, and no conclusion is presented as a reanalysis of newly acquired raw observations.