Water Quality Forecasting

Introduction

Here we want to forecast the spatio-temporal water quality in terms of the “power of hydrogen (pH)” value for the next day based on the input data, which is the historical data of other water measurement indices. The input data consists of daily samples for 36 sites, providing measurements related to pH values in Georgia, USA. The input features consist of 11 common indices including volume of dissolved oxygen, temperature, and specific conductance (see details in dataset). The output to predict is the measurement of 'pH, water, unfiltered, field, standard units (Median)'.

There are two major water systems to consider: one is centered on the city of Atlanta while the other is centered on the eastern coast of Georgia. This information indicates spatial depenency among different locations which are important to the forecast.

Processsed Data

Download link: [Dataset]

Data format: *.mat (use Matlab to open)

Data description:

Variable Name
 Type Size

Description
features array of strings 1*11 a list of water indices to measure
location_ids array of integer 37*1 IDs of the water stations
X_te array of matrices 1*282 test set input data: water indices for 282 contiguous dates until 2018-01-01
  • each element is a 37*11 matrix: 37 spatial locations by 11 features
X_tr array of matrices 1*423 training set input data: water indices for 423 contiguous dates from 2016-01-28
  • each element is a 37*11 matrix: 37 spatial locations by 11 features
Y_te array of matrices 37*282 test set output data: water quality for 37 locations in 282 contiguous dates until 2018-01-01
Y_tr array of matrices 37*423 training set output data: water quality for 37 locations in 423 contiguous dates from 2016-01-28
location_group array of cells 1*3 the groups of water stations, each group forms a connected spatial network (i.e., water system)

Data Source

This dataset is arranged and partly derived from the United States Geological Survey: [External Link]

Citation

To use these datasets, please cite the papers:

Liang Zhao, Olga Gkountouna, and Dieter Pfoser. 2019. Spatial Auto-regressive Dependency Interpretable Learning Based on Spatial Topological Constraints. ACM Trans. Spatial Algorithms Syst. 5, 3, Article 19 (August 2019), 28 pages. DOI:https://doi.org/10.1145/3339823

Acknowledgement

 

NSF 1755850 (sole-PI): "CRII: III: Interpretable Models for Spatio-Temporal Event Forecasting using Social Sensors", $174,990. 2018-2021, National Science Foundation.