Physically based distributed hydrologic models require geospatial and time-series data that take considerable time and effort in processing them into model inputs. Tools that automate and speed up input processing facilitate the application of these models. In this study, we developed a set of web-based data services called HydroDS to provide hydrologic data processing ‘software as a service.’ HydroDS provides functions for processing watershed, terrain, canopy, climate, and soil data. The services are accessed through a Python client library that facilitates developing simple but effective data processing workflows with Python. Evaluations of HydroDS by setting up the Utah Energy Balance and TOPNET models for multiple headwater watersheds in the Colorado River basin show that HydroDS reduces the input preparation time compared to manual processing. It also removes the requirements for software installation and maintenance by the user, and the Python workflows enhance reproducibility of hydrologic data processing and tracking of provenance.

基于物理的分布式水文模型需要地理空间和时间序列数据，这些数据需要花费大量时间和精力才能处理成模型输入。自动化和加速输入处理的工具有助于这些模型的应用。在这项研究中，我们开发了一套基于Web的数据服务，称为HydroDS，以提供水文数据处理“软件即服务”。HydroDS提供用于处理流域，地形，林冠，气候和土壤数据的功能。可通过Python客户端库访问这些服务，该客户端库有助于使用Python开发简单但有效的数据处理工作流。通过为科罗拉多河流域的多个源头流域建立犹他州能源平衡和TOPNET模型对HydroDS进行评估，结果表明与人工处理相比，HydroDS减少了投入准备时间。它还消除了用户对软件安装和维护的要求，并且Python工作流程增强了水文数据处理和源追踪的可重复性。