TITO V2.0.0

TITO Release v2.0: Real-time Precipitation Nowcasting & Multi-source Hydrologic Forecasting

This version of TITO expands the capabilities of the package by including machine learning-based nowcasting and support for multiple hydrologic forcing datasets.

What's included in this version

• Machine learning nowcasting:
  TITO now includes a standalone function to run a ConvLSTM algorithm for precipitation nowcasting. This function is independent of other repositories and solves issues in generating precipitation nowcast to fill the 4-hour latency gap of IMERG data.

• Precipitation data handling:
  Precipitation processing continues to rely on the GPM IMERG v07 Early Run product. The machine learning routines complement IMERG to provide near real-time nowcasts.

• EF5 hydrologic template improvements:
  Users can now run QPE and QPF from multiple sources through a "Long Range" function in the EF5 control file template. For this release, TITO supports:

  • GFS-based QPF: Download and process GFS data.
  • WRF-based QPF: Receive .nc format files and process them to convert them into EF5-compatible .tif format.
    For more details about the long range capability, visit the EF5 manual.

• End-to-end workflow:
  With this version, TITO can:

  1. Download and process IMERG to generate nowcasts and fill latency gaps.
  2. Download and utilize GFS/WRF for QPF.
  3. Run longer-range QPE and QPF runs from multiple sources through the "Long Range" function in the EF5 control file template.

Contributors

• @RobledoVD
• @humberva
• @namanmehta1208

We also thank early testers and collaborators for feedback that helped improve the workflow.

What's next

Future improvements may include:

• Extending support for additional high-resolution QPE and QPF products.
• Further enhancements to machine learning models.