| Autor: |
G. M. Southward, T. W. Sammis, P. Mott |
| Rok vydání: |
1994 |
| Předmět: |
|
| Zdroj: |
Applied Engineering in Agriculture. 10:41-44 |
| ISSN: |
1943-7838 |
| DOI: |
10.13031/2013.25825 |
| Popis: |
Automated weather collection systems are replacing climate instruments read by a person. However, data-loggers fail and data has to be estimated from adjacent climate stations. This article compares two linear regression methods for replacing data that was out of range or missing using the nearest climate stations. The first method uses three years of historical data and the second method uses the current climate year to estimate linear regression coefficients. The missing data in most cases can be estimated by using the current climate year and a good data block before the bad data. It is not necessary to wait to collect good data after the bad data block to estimate the linear regression coefficients used to replace the missing data based on good data from a nearby climate station. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
