A collection of high-resolution climate grid surfaces for land areas in Tasmania based on Climate Futures For Tasmania projection modelling. There are 152 climate products available that delineate temperature and rainfall parameters specific to crop growing requirements that form part of the enterprise suitability mapping program (refer here: https://dpipwe.tas.gov.au/agriculture/investing-in-irrigation/enterprise-suitability-toolkit/enterprise-suitability-maps). Broadly speaking these products include climate risk parameters including frost risk, heat risk and extreme rainfall risk as well as crop related indices including growing degree days and chill hours. Furthermore, mean monthly climate variables including mean monthly maximum/minimum air temperature and rainfall products are also produced. Refer here for dataset inventory: https://nrmdatalibrary.dpipwe.tas.gov.au/FactSheets/WfW/ListMapUserNotes/Inventory_DCM_Tas.pdf Climate Futures Tasmania (CFT) projections were incorporated into the modelling framework to simulate projected climate (according to the RCP 8.5 scenario) for years 2030 and 2050. These projections were downscaled, and bias corrected to a spatial grid resolution of 30m. Also, note that these outputs relate to the baseline climate maps defined here: https://www.thelist.tas.gov.au/app/content/data/geo-meta-data-record?detailRecordUID=ba62f124-5906-4471-a01c-9b57b6142055 All products can be accessed via Web Map Service: https://spatial.dpipwe.tas.gov.au/naturalassets/Climate/wms Or viewed in the following Web Map application: https://arcg.is/vaHDG

This project set out to incorporate and downscale six regional climate models (RCMs) produced from the Climate Futures for Tasmania (CFT) project (refer here: https://climatefutures.org.au/projects/climate-futures-tasmania/) to a resolution fit for local crop suitability mapping (spatial resolution of 30m) in Tasmania, Australia and to complement the baseline climate map series (refer here: https://www.thelist.tas.gov.au/app/content/data/geo-meta-data-record?detailRecordUID=ba62f124-5906-4471-a01c-9b57b6142055). The climate models were in accordance to the Representative Concentration Pathways scenario 8.5, i.e. for a future climate with little curbing of CO2 emissions (RCP 8.5.) Refer to the following link for a more formal definition of the Representative Concentration Pathways scenarios: https://climatechangeinaustralia.gov.au/en/overview/understanding-projections/ The downscaling procedure used regression trees and covariate data to disaggregate each RCM to over 500 baseline sites which were then bias adjusted to historic temperature and rainfall data (based on DPIPWE and Bureau of Meteorology weather station and temperature logger recordings for the baseline period: 1998-2017). The bias-adjustment method incorporated a non-parametric quantile mapping technique with empirical quantiles transformation (Gudmundsson 2016). Based on this analysis, climatic variables such as frost risk, heat risk, growing degree days (GDD), growing season temperature (GST), rainfall and chill hours were then determined for two projection periods: 2011-2030 and 2031-2050. This occurred by firstly calculating the aforementioned climate variables at the baseline sites (based on the bias adjusted data) in accordance to their specific rule parameter (refer to https://nrmdatalibrary.dpipwe.tas.gov.au/FactSheets/WfW/ListMapUserNotes/Inventory_DCM_Tas.pdf) to produce the projected climate estimates. Secondly, spatial interpolating these values to produce a continuous raster surface representing each climate variable. Regression trees and thin plate splines was used to interpolate the climate variables. Since each climate variable are based separately from each of the 6 RCMs, each of the resulting surfaces was collectively placed in a six member ensemble stack to derive the mean value. The resultant output became the final representative climate layer in line with the specified rule parameter and projection year (as defined in the link above). A document that describes each dataset can be accessed here: https://nrmdatalibrary.dpipwe.tas.gov.au/FactSheets/WfW/ListMapUserNotes/Inventory_DCM_Tas.pdf All outputs produced from this modelling are available for use in a GIS via a Web Map Service (WMS): https://spatial.dpipwe.tas.gov.au/naturalassets/Climate/wms References: Gudmundsson, L. (2016) R package 'qmap': Statistical Transformations for Post-Processing Climate Model Output. R package version 1.0-4. https://cran.r-project.org/web/packages/qmap/qmap.pdf

Position Accuracy: 30m. Attribute Accuracy: Preliminary assessment of the downscaling method revealed that the climatic variables inherited the general spatial trends of their equivalent coarse scale counterpart provided by the raw CFT data. This revealed that the correlation coefficient (R2) between the downscaled estimates and their corresponding raw CFT grid cell were always above >0.8, signifying strong agreement between raw CFT data and downscaled values. From this, it is assumed that resulting spatial outputs are within acceptable predictions ranges emanating from the downscaling and bias correction processes employed. Please note that results from a comprehensive uncertainty analysis will be provided in due course when finalised.