The Addition of Temperature to the TSS-RESTREND Methodology Significantly Improves the Detection of Dryland Degradation

Autor: Arden L. Burrell, Yi Liu, Jason P. Evans
Rok vydání: 2019
Předmět:
Zdroj: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing. 12:2342-2348
ISSN: 2151-1535
1939-1404
DOI: 10.1109/jstars.2019.2906466
Popis: Cold drylands make up 20% of the world's water-limited regions. This paper presents a modification to the Time Series Segmented-RESidual TRENDs (TSS-RESTRENDs) method that allows for the use of temperature as an additional explanatory variable along with precipitation. TSS-RESTREND was performed over Mongolia, both with and without temperature. The addition of temperature reduced the number of pixels that fail the significance tests built into the TSS-RESTREND method from 17% to below 5%. It also improved the statistical confidence in almost all areas. Furthermore, the direction of change is consistent with previous findings that looked at vegetation trends over the same study region. When applied to all of the world's drylands, the inclusion of temperature improved the fit of the vegetation–climate relationship that underpins TSS-RESTREND in 98.8% of areas. The largest improvements to the fit were observed in both the cold drylands of central Asia and North America and the hot drylands of southern Australia. Including temperature also reduced the fraction of global vegetation change that could be attributed to neither climate nor land use from 25.5% to 15.5%.
Databáze: OpenAIRE