Climate Victoria: Minimum Temperature (3DS-T; 9 second, approx. 250 m)

Daily (1981-2019), monthly (1981-2019) and monthly mean (1981-2010) surfaces of minimum temperature (approx. 1.2 m from ground) across Victoria at a spatial resolution of 9 seconds (approx. 250 m). Surfaces are developed using trivariate splines (latitude, longitude and elevation) with partial dependence upon a topographic index of relative elevation.\nLineage: A) Data modelling:\n1. Weather station observations collected by the Australian Bureau of Meteorology were obtained via the SILO patched point dataset (https://data.qld.gov.au/dataset/silo-patched-point-datasets-for-queensland), followed by the removal of all interpolated records.\n2. Climate normals representing the 1981-2010 reference period were calculated for each weather station. A regression patching procedure (Hopkinson et al. 2012) was used to correct for biases arising due to differences in record length where possible.\n3. Climate normals for each month were interpolated using trivariate splines (latitude, longitude and elevation as spline variables) with partial dependence upon a topographic index of relative elevation. All models were fit and interpolated using ANUSPLIN 4.4 (Hutchinson & Xu 2013).\n4. Daily anomalies were calculated by subtracting daily observations from climate normals and interpolated with full spline dependence upon latitude and longitude\n5. Interpolated anomalies were added to interpolated climate normals to obtain the final daily surfaces. \n6. Monthly surfaces are calculated as an aggregation of the daily product.\nB) Spatial data inputs:\n1. Fenner School of Environment and Society and Geoscience Australia. 2008. GEODATA 9 Second Digital Elevation Model (DEM-9S) Version 3.\nC) Model performance (3DS-T):\nAccuracy assessment was conducted with leave-one-out cross validation. \nMean monthly minimum temperature RMSE = 0.80 °C\nDaily minimum temperature RMSE = 1.73 °C\n\nPlease refer to the linked manuscript for further details.\n