This README file is associated with data and analyses from publication "Quantifying thermal exposure for migratory riverine species: phenology of Chinook salmon populations predicts thermal stress" by FitzGerald et al., Global Change Biology 2020. The stream temperature modelling results from the spatial stream network (SSN) model are attached as 8 shapefile feature classes (.shp), with each corresponding to a sub-regional watershed, in projection GCS_NAD_83 to be used in mapping software (e.g. ESRI ArcMap, QGIS). Fields are as follows: 'FID', Shape', and 'OBSPRED_ID' are generic descriptors and identifiers. 'FTYPE' and 'WATERBODY' were not used in data analysis, but were used to eliminate reaches with unsuitable predictions. 'COMID' and 'GNIS_NAME' are unique identifiers and stream names outputted from NHDPlus. The following spatial covariates were used in the SSN: 'TAILWATER', 'ELEV', 'CANOPY', 'SLOPE', 'PRECIP', 'CUMDRAING', 'Y_COORD', 'NLCD11PC', 'GLACIER', and 'BFI' (see text for descriptions and units of each covariate). Two temporal covariates, air temperature (degrees C) and flow (m^3/s), were included for each month for each of 8 sub-regional watersheds. They are labeled as 'Air_Month' and 'Flow_Month' (e.g. 'Air_Jan' and 'Flow_Jan' indicate the air temperature and flow for January). Output results are: the mean monthly stream temperature for each stream segment from the period 2002-2011, in the format 'Water_Month'; and the prediction standard error ('Water_MonthSE').