当前位置: X-MOL 学术Atmos. Ocean › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
On the Discretization of Richards Equation in Canadian Land Surface Models
Atmosphere-Ocean ( IF 1.2 ) Pub Date : 2022-07-14 , DOI: 10.1080/07055900.2022.2096558
Murray D. MacKay 1 , Gesa Meyer 2 , Joe R. Melton 2
Affiliation  

ABSTRACT

Both the Canadian Land Surface Scheme (CLASS) and the Soil, Vegetation, and Snow (SVS) land surface models employ a potentially problematic discretization of Richards equation for unsaturated vertical flow in the soil column. It is shown here that this discretization will always overestimate the vertical moisture gradient compared to a better-constructed first-order scheme, which under some circumstances could lead to erroneous moisture drawdown. The problem stems from an interpolation calculation on the irregularly spaced grids traditionally used by these models. While vanishing on uniform grids, this numerical error progressively worsens with increasing layer thickness differences. In this brief technical note a systematic method for developing first and higher order schemes on irregular, staggered grids is presented. To demonstrate the potential impact of the new first – order scheme, multi-year simulations of five FLUXNET sites are presented and discussed. A dramatic improvement in first layer soil moisture is found for two of the sites, which contributes to potentially significant differences in evapotranspiration. Higher order schemes are also possible but must be constructed carefully, in concert with a judicious choice of soil layer spacing in order to minimize discretization error. Given the extensive use of CLASS and SVS in Canadian environmental prediction systems, and the freedom with which users can specify soil layer thicknesses, it is recommended that modellers consider this issue carefully in their applications.



中文翻译:

关于加拿大地表模型中 Richards 方程的离散化

摘要

加拿大地表方案 (CLASS) 和土壤、植被和雪 (SVS) 地表模型都对土柱中的非饱和垂直流采用了可能存在问题的 Richards 方程离散化。此处显示,与更好构造的一阶方案相比,这种离散化总是会高估垂直水分梯度,这在某些情况下可能导致错误的水分下降。问题源于对这些模型传统上使用的不规则间隔网格的插值计算。当在均匀的网格上消失时,这种数值误差会随着层厚差异的增加而逐渐恶化。在这篇简短的技术说明中,介绍了一种在不规则、交错网格上开发一阶和高阶方案的系统方法。为了证明新的一阶方案的潜在影响,提出并讨论了五个 FLUXNET 站点的多年模拟。在两个地点发现第一层土壤水分显着改善,这有助于蒸发蒸腾的潜在显着差异。高阶方案也是可能的,但必须仔细构建,与土壤层间距的明智选择相一致,以尽量减少离散化误差。鉴于 CLASS 和 SVS 在加拿大环境预测系统中的广泛使用,以及用户可以自由指定土壤层厚度,建议建模者在他们的应用中仔细考虑这个问题。在两个地点发现第一层土壤水分显着改善,这有助于蒸发蒸腾的潜在显着差异。高阶方案也是可能的,但必须仔细构建,与土壤层间距的明智选择相一致,以尽量减少离散化误差。鉴于 CLASS 和 SVS 在加拿大环境预测系统中的广泛使用,以及用户可以自由指定土壤层厚度,建议建模者在他们的应用中仔细考虑这个问题。在两个地点发现第一层土壤水分显着改善,这有助于蒸发蒸腾的潜在显着差异。高阶方案也是可能的,但必须仔细构建,与土壤层间距的明智选择相一致,以尽量减少离散化误差。鉴于 CLASS 和 SVS 在加拿大环境预测系统中的广泛使用,以及用户可以自由指定土壤层厚度,建议建模者在他们的应用中仔细考虑这个问题。与土壤层间距的明智选择相一致,以尽量减少离散化误差。鉴于 CLASS 和 SVS 在加拿大环境预测系统中的广泛使用,以及用户可以自由指定土壤层厚度,建议建模者在他们的应用中仔细考虑这个问题。与土壤层间距的明智选择相一致,以尽量减少离散化误差。鉴于 CLASS 和 SVS 在加拿大环境预测系统中的广泛使用,以及用户可以自由指定土壤层厚度,建议建模者在他们的应用中仔细考虑这个问题。

更新日期:2022-07-14
down
wechat
bug