Subject: Marine Geospatial Ecology Tools (MGET) help
Text archives
From: | "Jason Roberts" <> |
---|---|
To: | "'Jiayi Wang'" <> |
Cc: | <> |
Subject: | RE: [mget-help] A question about Auqa-MODIS chlorophyll data |
Date: | Sat, 1 Dec 2012 13:12:35 -0500 |
Jiayi, My guess is that either 1) the chlorophyll HDF variable does not have a scaling equation, but that you applied one, or 2) the chlorophyll HDF variable does have a scaling equation but that you did not apply it. It used to be that ocean color HDFs from NASA were distributed as integer variables, and to get the actual floating-point values you had to apply a scaling equation to those integers. NASA did this because the integers would only require 16 bits of space per pixel, compared to the 32 bits that would be required if they distributed the data in floating-point format. Users that were not aware of this optimization by NASA would often encounter problem #2. In my original tutorial about converting HDFs (here), I described how to find the scaling equation in the chlorophyll variable and apply it to the integers to compute the floating-point values. But a couple of years ago NASA started distributing HDFs with floating-point values instead. I guess they discovered that using floating-point format did not result in too much increase in file size, given that they were compressing the HDF files anyway. When this happened, a person who blindly followed by tutorial would encounter problem #1: they assumed the chlorophyll values were unscaled integers, but in fact they were scaled floating-point values. They applied the scaling equation a second time, resulting in bizarre chlorophyll values. In any case, for people needing chlorophyll data from NASA in ArcGIS-compatible format, I now suggest you do not download the HDFs and convert them manually. Instead, inside the MGET toolbox, go into Data Products à NASA GSFC OceanColor Group and use the tool inside there called Created Rasters for NASA OceanColor L3 SMI Product. This tool automates all the steps of downloading the HDFs and converting them. Please try that tool and let me know if it works for you. Also, even though my HDF conversion tutorial uses the NOAA NODC 4km AVHRR Pathfinder 5.0 SST data as an example, I actually recommend that people not download those manually either. There is a similar automated tool for downloading and converting that data as well. Best regards, Jason From: Jiayi Wang [mailto:] Hi, I downloaded the Aqua-MODIS chlorophyll concentration data (HDF) and converted it to ArcGIS raster following your instruction. I converted 4 HDF files and everything looked great in ArcGIS, but then I found the value range of 4 maps are all approximately from 0 to 100. Could you please tell me why they have almost the same value range? Thanks! Jiayi |