An inflection point-based method for estimating metrics of mangrove phenology combining climatic factors and Landsat NDVI time series

Abstract

The present research evaluated the prospects of utilizing rainfall and temperature combined with Landsat-8 derived HANTS (Harmonic Analysis of Time Series) reconstructed NDVI for estimating the metrics of the mangrove phenology. The selected period of the study was from 2013 to 2020 for the Pichavaram mangroves of Tamil Nadu. The NDVI and ERA5 (ECMWF Re-Analysis) datasets of rainfall and temperature were the input datasets for developing the new algorithm. The ‘z-score sum’ provided a measure of the cumulative impact of rainfall and temperature, displaying its most negative value coinciding with the peak positive value of the NDVI time series datasets. The algorithm developed for phenological metrics estimation identified the common inflection points of the z-score sum and NDVI curves. The temporal analysis of metrics revealed the average Length of Season (LoS) as 230 days. The metrics also identified the drought year 2016 with the shortest LoS and the least Gross Primary Productivity (GPP) values. The analysis showed the influences of the preceding year’s monsoon rainfall on the GPP values of the later part of the phenological cycle. The temperatures during the days of PoS were found to be the optimum temperature for the growth of mangroves. HIGHLIGHTS The combined influence of rainfall and temperature on mangrove phenology is more significant than individual influences.; The z-score sum of rainfall and temperature exhibited an inverse relationship with NDVI values.; GPP values of mangroves and estimated phenological metrics showed good correspondence.; The estimated phenological metrics captured the influences of drought and abundant rainfall on mangrove phenology.;

Keywords

• gross primary productivity
• harmonic analysis of time series
• rainfall
• temperature
• z-score