Sequential Imputation of Missing Spatio-Temporal Precipitation Data Using Random Forests

Abstract

Meteorological records, including precipitation, commonly have missing values. Accurate imputation of missing precipitation values is challenging, however, because precipitation exhibits a high degree of spatial and temporal variability. Data-driven spatial interpolation of meteorological records is an increasingly popular approach in which missing values at a target station are imputed using synchronous data from reference stations. The success of spatial interpolation depends on whether precipitation records at the target station are strongly correlated with precipitation records at reference stations. However, the need for reference stations to have complete datasets implies that stations with incomplete records, even though strongly correlated with the target station, are excluded. To address this limitation, we develop a new sequential imputation algorithm for imputing missing values in spatio-temporal daily precipitation records. We demonstrate the benefits of sequential imputation by incorporating it within a spatial interpolation based on a Random Forest technique. Results show that for reliable imputation, having a few strongly correlated references is more effective than having a larger number of weakly correlated references. Further, we observe that sequential imputation becomes more beneficial as the number of stations with incomplete records increases. Overall, we present a new approach for imputing missing precipitation data which may also apply to other meteorological variables.

One year of attenuation data from a commercial dual-polarized duplex microwave link with concurrent disdrometer, rain gauge and weather observations

“…event on 5 August 2011). How this can be solved by "inputting" missing observations based on signals from nearby sensors (Mital et al, 2020), remains to be seen.…”

Challenging Problems of Quality Assurance and Quality Control (QA/QC) of Meteorological Time Series Data

“…We have conducted imputation of these gaps. The application of multivariate imputation of missing values for groundwater levels can be found in the paper by Dwivedi et al, 2021, and for precipitation--in the paper by Mital et al, 2020. 5.2.2. Imputation using the library ImputeTS Imputation was conducted using the na_seadec () function (Seasonally Decomposed Missing Value Imputation) of the ImputeTS package in R, with a the time series frequency of 365, and using the algorithm "interpolation."…”

Reliable modeling and prediction of precipitation & radiation for mountainous hydrology

“…AI-based surrogate atmospheric process modeling can also be of great utility for assessing the value added by additional observations, and provides an alternative pathway that avoids structural uncertainties in TPMs. Surrogate models can readily incorporate rich datasets, such as from SAIL [ Chen et al , 2020;Liu et al , 2020;Mital et al , 2020], and can quickly discern obvious and non-obvious relationships that impact precipitation and radiation without having to first sleuth out WRF model errors. Surrogate models, such as deep neural networks, have already been shown to discern these relationships [ Weber et al , 2020].…”