Climate Data Record (CDR) Program
Climate Algorithm Theoretical Basis Document (C-ATBD)
Precipitation – PERSIANN-CDR
CDR Program Document Number: CDRP-ATBD-0286
Configuration Item Number: 01B-16
Revision 2 / March 21, 2014
CDR Program PERSIANN C-ATBD CDRP-ATBD-0286
Rev. 2 05/12/2014
TABLE of CONTENTS
1. INTRODUCTION............................................................... 7
1.1 Purpose............................................................................... 7
1.2 Definitions............................................................. 7
1.3 Document Maintenance................................................8
2. OBSERVING SYSTEMS OVERVIEW........................ 9
2.1 Products Generated .....................................................9
2.2 Instrument Characteristics.............................................9
3. ALGORITHM DESCRIPTION......................................10
3.1 Algorithm Overview ................................................ 10
3.2 Processing Outline....................................................... 10
3.3 Algorithm Input.................................................... 11
3.3.1 Primary Sensor Data .................................................11
3.3.2 Ancillary Data................................................11
3.3.3 Derived Data ...........................................................12
3.3.4 Forward Models.............................................................12
3.4 Theoretical Description ..................................................... 12
3.4.1 Physical and Mathematical Description............................12
3.4.2 Data Merging Strategy.....................................................15
3.4.3 Numerical Strategy .................................................15
3.4.4 Calculations...................................................................15
3.4.5 Look-Up Table Description........................................16
3.4.6 Parameterization ............................................................16
3.4.7 Algorithm Output................................................................16
4. TEST DATASETS AND OUTPUTS....................................17
4.1 Test Input Datasets .................................. 17
4.2 Test Output Analysis ...................................... 18
4.2.1 Reproducibility................................................................18
4.2.2 Precision and Accuracy ....................................22
4.2.3 Error Budget.....................................................22
5. PRACTICAL CONSIDERATIONS.................................23
5.1 Numerical Computation Considerations................................ 23
5.2 Programming and Procedural Considerations ...................... 23
5.3 Quality Assessment and Diagnostics ............................ 23
5.4 Exception Handling ............................................................... 23
5.5 Algorithm Validation......................................................... 23
5.6 Processing Environment and Resources................................. 24
6. ASSUMPTIONS AND LIMITATIONS .......................25
6.1 Algorithm Performance............................... 25
6.2 Sensor Performance............................................... 25
7. FUTURE ENHANCEMENTS.......................................26
8. REFERENCES......................................................................27
APPENDIX A. ACRONYMS AND ABBREVIATIONS................29
1. Introduction
1.1 Purpose
The purpose of this document is to describe the algorithm submitted to the National Climatic Data Center (NCDC) by Soroosh Sorooshian at the Center for Hydrometeorology and Remote Sensing, at University of California, Irvine, that will be used to create the PERSIANN-CDR (Precipitation Estimation from Remotely Sensed Information using an Artificial Neural Network-Climate Data Record), using the long wave infrared images from geosynchronous satellites. The actual algorithm is defined by the computer program (code) that accompanies this document, and thus the intent here is to provide a guide to understanding that algorithm, from both a scientific perspective and in order to assist a software engineer or end-user performing an evaluation of the code.
1.2 Definitions
PERSIANN rain rate estimates are generated as a 0.25 degree resolution product that is then calibrated to the monthly 2.5 degree merged in-situ and satellite product produced by the Global Precipitation Climatology Project. GridSat-B1 IRWIN data are used as input to the PERSIANN model at 0.25 degree resolution and a 3-hourly time step. The output from the PERSIANN model (before bias correction) is called PERSIANN-B1. A threshold (thd) value needs to be applied to the 3-hourly PERSIANN-B1 rain rate estimates to filter out noisy pixels. These noisy pixels are generally associated with pixels where the rain rate is “zero” but the Neural Network model estimates a very small nonzero value. While the resulting noisy pixels may not affect the adjustment process considerably, they can lead to a very large number of "rainy" days (rain rate > 0 mm/day). The PERSIANN-B1 data will be accumulated to monthly and 2.5 degrees for GPCP bias weight calculation (w). We note that in some locations, such as high latitudes and in dry regions with very low rainfall values, w can become large. This can lead to unreasonably large daily rainfalls in finer resolution. In order to prevent such cases, we applied a cap for the maximum weight. In order to find the best combination of thd and maximum w, an optimization model was developed with the objective of finding the combination which gives the minimum Mean Absolute Error (MAE) between GPCP-1DD and PERSIANN-CDR (up-scaled to 1o). The results show that thd = 0.1 and maximum w = 20 is perhaps the best combination.
Full Text
Download Click here
Read and download Click here
ليست هناك تعليقات:
إرسال تعليق