Abstract
We compare microwave GPS and optical-based remote sensing observations of the vegetation response to a recent drought in California, USA. The microwave data are based on reflected GPS signals that were collected by a geodetic network. These data are sensitive to temporal variations in vegetation water content and are made available via the Normalized Microwave Reflection Index (NMRI). NMRI data are complementary to information of plant greenness provided by the Normalized Difference Vegetation Index (NDVI). NMRI data from 146 sites in California are compared to collocated NDVI observations, over the interval of 2007–2016. This period includes a severe, three-year drought (2012–2014). We quantify the seasonal variations in vegetation state by calculating a series of phenology metrics at each site, using both NMRI and NDVI. We examine how the phenology metrics vary from year-to-year, as related to the observed fluctuations in accumulated precipitation. The amplitude of seasonal vegetation growth exhibits the greatest sensitivity to prior accumulated precipitation. Above-normal precipitation from 4 to 12 months before peak growth yields a stronger seasonal growth pulse, and vice versa. The amplitude of seasonal growth, as determined from NDVI, varies linearly with precipitation during dry years, but is largely insensitive to precipitation amount in years with above-normal precipitation. In contrast, the amplitude of seasonal growth from NMRI varies approximately linearly with precipitation across the entire range of conditions observed. The length of season is positively correlated with prior accumulated precipitation, more strongly with NDVI than NMRI. The recovery from drought was similar for a one-year (2007) and the more severe three-year drought (2012–2014). In both cases, the amplitude of growth returned to typical values in the first year with near-normal precipitation. Growing season length, only based on NDVI, was greatly reduced in 2014, the driest and final year of the three-year California drought.
Highlights
Monitoring and characterizing drought depends upon the application [1]
We summarize the magnitude of prior precipitation anomalies for any month (e.g., April 2014) as Percent of Normal Precipitation (PNP), which was calculated over intervals from 1 to 12 months in duration
Prior to considering the response to drought, we first summarize differences and similarities between the phenology metrics calculated from Normalized Microwave Reflection Index (NMRI) and Normalized Difference Vegetation Index (NDVI)
Summary
Monitoring and characterizing drought depends upon the application [1]. Ecological, agricultural, meteorological, and hydrological drought are all quantified with different data. The measured response of vegetation to drought depends on the data and metrics that are used to gauge changes in vegetation amount or function. The Normalized Difference Vegetation Index (NDVI), which is a combination of reflectance in the near infrared and red bands, is a measure of the capacity of vegetation to absorb photosynthetically active radiation [7]. NDVI is more widely used to characterize the temporal variations in vegetation state, for example, in climate and hydrologic models, due to the long period of record made possible by combining data from AVHRR and MODIS [9]. Most optical remote sensing metrics are sensitive to plant greenness, metrics such as the Normalized Difference Water Index (NDWI) provides a measure of water in vegetation canopies [10]. More useful information on vegetation water content may be gained from hyperspectral measurements, but these data are currently limited in their spatial coverage and repeat time
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
