LANLoad NEEPP: Landscape Assessment of Nutrient Loading to Waterbodies (LANLoad) in the Northern Everglades and Estuaries Protection Program (NEEPP) region

LANLoad is a geospatial screening tool designed to facilitate water quality management decisions. It provides an estimate of the relative likelihood that nutrient inputs applied at specific locations on land will impact water quality. LANLoad is based solely on physical characteristics and may be used independently or with other relevant datasets.

LANLoad NEEPP is available as a single comprehensive file "LANLoad_NEEPP_Overall" and as subsets corresponding to intersections between NEEPP and 15 FL counties.

The datasets consist of cells (10m x 10m) ranked to reflect the likelihood that nutrients applied to a given terrestrial location will reach a downgradient surface waterbody. Possible ranks range from 1 to 9 with values increasing as the likelihood of nutrient transport to downgradient surface waterbodies increases. Ranks are based on 6 physical landscape parameters selected by Subject Matter Experts (SMEs) who also assigned relative weights to each parameter using the Analytical Hierarchy Process (AHP). During this exercise, the location considered by SMEs was the pilot study area, St Lucie County, FL, and the focal nutrient source was Onsite Sewage and Treatment Disposal Systems (OSTDS). Despite the original focus on OSTDS, LANLoad NEEPP can be used to gauge the likelihood of nutrient transport to surface waterbodies from other, similar, nutrient sources. The resulting AHP model demonstrated high internal consistency (Consistency Ratio: 0.01) and resulted in the following parameter weights, in order of importance:

• Distance to Waterbody, 30.0%;
• Depth to Water, 21.6%;
• Hydraulic Conductivity, 20.7%;
• Potential for Flooding, 10.9%;
• Slope, 9.8%; and
• Surficial Karstic Deposits, 7.0%.

Geospatial datasets representative of these parameters were acquired (2024) and combined using a weighted overlay to produce LANLoad NEEPP. Details are available in a report (link below) and publication (in prep as of Jan 2026)

LANLoad NEEPP performance was evaluated at multiple locations (selected via a random stratified process) within NEEPP by classifying LANLoad ranks less than or equal to 4 as “lower” and those more than or equal to 6 as “higher”. Then, two assessment methods were applied, both conducted blind:

1) SME Review: SMEs were provided with input datasets corresponding to 30 locations and asked to assign a classification of lower or higher. There was 92 % consistency between classifications assigned by LANLoad NEEPP and those assigned by SMEs.

2) Numerical modeling: Using ArcNLET-Py, nutrient loading to surface waters from uniform inputs was modeled in 10 locations, each containing 50 model points. Classifications assigned by LANLoad were 100% consistent with those assigned through ArcNLET-Py model results, i.e., locations classified by LANload as “higher” also had the highest ArcNLET-Py modeled nutrient loads while those classified as “lower” had the lowest modeled nutrient loads.

Contact: Kai Rains – [e-mail]

LANLoad是一款专为辅助水质管理决策而设计的地理空间筛查工具，可估算陆地特定位置施用的养分输入影响水质的相对可能性。LANLoad仅基于物理特征构建，可独立使用，也可与其他相关数据集结合使用。

LANLoad NEEPP 以单一综合文件"LANLoad_NEEPP_Overall"形式发布，同时也提供与NEEPP及15个佛罗里达州（FL）下辖县的交集对应的子集数据。

该数据集由10米×10米的栅格单元组成，各单元被赋予等级以反映施用于特定陆地区域的养分抵达下游地表水体的可能性。等级取值范围为1至9，数值越高，养分向下游地表水体运移的可能性就越大。

等级评定基于6项由领域专家（Subject Matter Experts, SMEs）遴选的物理景观参数，领域专家还通过层次分析法（Analytical Hierarchy Process, AHP）为各参数赋予了相对权重。本次建模所考量的试点研究区域为佛罗里达州圣露西县（St Lucie County, FL），核心养分来源为就地污水处理系统（Onsite Sewage and Treatment Disposal Systems, OSTDS）。尽管最初的研究聚焦于就地污水处理系统，但LANLoad NEEPP亦可用于评估其他类似养分来源向地表水体运移养分的可能性。最终构建的AHP模型内部一致性极高（一致性比率：0.01），按重要性排序的各参数权重如下：
• 距水体距离：30.0%；
• 地下水位埋深：21.6%；
• 导水率：20.7%；
• 洪涝潜势：10.9%；
• 坡度：9.8%；
• 表层喀斯特沉积：7.0%。

本次研究所用的表征上述参数的地理空间数据集采集于2024年，并通过加权叠加分析整合得到LANLoad NEEPP。详细信息可参阅下述报告及2026年1月筹备发表的学术论文。

LANLoad NEEPP的性能在NEEPP范围内的多个区域（通过随机分层抽样流程选取）进行了评估。评估时将LANLoad等级≤4的划分为“低风险”，等级≥6的划分为“高风险”。随后采用两种盲法评估方法：
1. 领域专家评审：向领域专家提供对应30个采样点的输入数据集，要求其将对应区域划分为“低风险”或“高风险”。LANLoad NEEPP的分类结果与领域专家的分类结果一致性达92%。
2. 数值模拟：借助ArcNLET-Py工具，对10个采样区域（每个区域包含50个模拟点）中均匀输入的养分向地表水体的负荷量进行建模。LANLoad的分类结果与ArcNLET-Py模拟结果完全一致：即LANLoad标记为“高风险”的区域，其ArcNLET-Py模拟养分负荷量亦最高；而被标记为“低风险”的区域，模拟养分负荷量则最低。

联系方式：Kai Rains – [e-mail]